For management activities, especially in the decision-making process, the most useful models are those that are expressed in words or formulas, algorithms and other mathematical means.
The basis of management based on loyalty was laid in 1908 by Harvard professor J. Royce. He is the author of the book "Philosophy of Loyalty", where the concept of "loyalty" is scientifically defined for the first time.
Within the framework of the proposed verbal model, business loyalty is considered from the point of view of three independent basic aspects: consumer loyalty, employee loyalty and investor loyalty.
Each time, the word "loyalty" means something different. Meskon M.Kh., Albert M., Hedouri F. Fundamentals of Management / Per. from English. - M., 2002. - P. 456 .:
commitment (from the point of view of buyers),
Integrity (from the point of view of employees),
· Mutual trust, respect and support (from the point of view of investors).
But, despite the pronounced components, this system should be considered only as a whole, since it is impossible to create loyal customers without paying attention to employee loyalty, or to cultivate employee loyalty without due attention to investor loyalty. None of the parts can exist separately from the other two, but all three together allow the organization to reach unprecedented heights in development.
It must be clearly understood that loyalty-based management is primarily focused on people. First of all, it is people and their role in business that are considered here. It is more a model of motivation and behavior than marketing, financial or production development. Only secondarily does loyalty-based management generalize people into more abstract categories and manage technical processes.
As practice shows, people are always more willing to work for an organization that has a purpose of service than for an organization that exists only to "make money". Therefore, people willingly work in the church or in public organizations.
Managers who want to successfully use the loyalty effect management model should not consider profit as a primary goal, but as a necessary element for the well-being and survival of the three components of every business system: customers, employees, and investors. Even at the beginning of the twentieth century. G. Ford said that “an organization cannot work without profit, ... otherwise it will die. But to create an organization only for the sake of profit ... means to lead it to certain death, since it will not have an incentive to exist ” Drucker P.F. Tasks of management in the XXI century. - M., 2001. - S.523 ..
The basis of the loyalty model under consideration is not profit, but the attraction of additional customers, a process that consciously or unconsciously underlies most successful organizations. Creating a target number of buyers permeates all areas of a company's business. The forces that govern the relationship between customers, employees, and investors are called the forces of loyalty. The measure of success is whether customers come back to buy more, or whether they go somewhere else, ie. whether they are loyal.
As a reason, loyalty initiates several economic effects that affect the entire business system in the following way Repin V.V., Eliferov V.G. Process approach to management: Modeling of business processes. - M., 2005. - 2nd ed. - P.245 .:
1. Profit and market share grow when the most promising buyers cover the entire range of the company's activities, creating good things about it public opinion and keep shopping. Due to the large and high-quality offer, the company can afford to be more picky when choosing new customers and focus on more profitable and potentially loyal projects to attract them, further stimulating its long-term growth.
2. Long-term growth allows the firm to attract and retain the best employees. Consistently maintaining a target number of buyers increases employee loyalty, giving them a sense of pride and job satisfaction. Further, in the process of interaction, regular employees learn more about their regular customers, in particular, how to better serve them so that the volume of purchases grows. This increasing volume of sales spurs both customer loyalty and employee loyalty.
3. Loyal employees in long term learn to reduce costs and improve the quality of work (learning effect). The organization can use this extra productivity to expand the reward system, to buy the best equipment and learning. All this, in turn, will spur employee productivity, reward growth and, consequently, loyalty.
4. This productivity spiral provides a cost advantage that is very difficult to replicate for purely competitive organizations. Long-term cost advantages, coupled with a steady growth in the number of loyal customers, bring profits that are very attractive to investors. This, in turn, enhances the company's ability to attract and retain the "right" investors.
5. Loyal investors act like partners. They stabilize the system, lower the cost of raising capital, and ensure that diverted cash flows are put back into the business as an investment. This strengthens the organization and increases its productive capacity.
Without a doubt, each organization is unique, but still, to one degree or another, its profit indicators will fit into general model economic effects derived from the persistence or loyalty of customers. Among them, it is worth noting the following Meskon M.Kh., Albert M., Hedouri F. Fundamentals of Management / Per. from English. - M., 2002. - S. 358 .:
basic profit (the price paid by newly appeared buyers exceeds the cost of the organization to create a product);
revenue growth (as a rule, if the buyer is satisfied with the parameters of the product, he is inclined to increase the volume of purchases over time);
Savings costs (close familiarity with the organization's products reduces the dependence of buyers on its employees for information and advice);
Reviews (customers satisfied with the level of service recommend the organization to their friends and acquaintances);
additional price (regular customers who cooperate with the organization long enough to explore all of its products and services receive disproportionately more from continuing the relationship and do not need additional discounts or promotions).
To assess the true long-term loyalty potential of a customer or group of customers, it is necessary to know their propensity to exhibit consistency. So some buyers will defect to a competitor for a 2% discount, while others will remain at a 20% price difference. The amount of effort it takes to lure different types of customers is called the loyalty ratio. In some organizations, the history of development or the behavior of customers in individual segments is used to assess loyalty coefficients Repin V.V., Eliferov V.G. Process approach to management: Business process modeling. - M., 2005. - 2nd ed. - P.232.. In others, especially those whose future is weakly connected with the past, they try to find out by data analysis methods how big the discount should be so that buyers go to their organization. But despite all the measurement challenges, using a loyalty metric allows organizations to identify customer retention and implement sound practices proven in one department throughout the organization.
Development of measurement, analysis and control systems cash flows obtained from loyalty can lead the organization to investments that will further ensure the growth of the number of customers and the organization as a whole.
So, the loyalty model is substantiated in detail at the verbal level. This justification mentioned mathematical and computer support. However, they are not required to make initial decisions.
With a more thorough analysis of the situation, verbal models, as a rule, are not enough. It is necessary to use sufficiently complex mathematical models. Thus, when making decisions in management production systems Kuzin B.I., Yuriev V.N., Shakhdinarov G.M. are used. Methods and models of firm management: Proc. for universities. - SPb., 2001. - P.327.
models technological processes(primarily models of control and management);
Models for ensuring product quality (in particular, models for assessing and controlling reliability);
queuing models;
Inventory management models (logistics models);
Simulation and econometric models of the enterprise as a whole, etc.
- improving the "as it should be" model. Business process modeling is not limited to creating a “how it should be” model. Each of the processes continues to change and improve along the way, so process models should be regularly reviewed and improved. This stage of modeling is associated with continuous improvement of processes and improvement of the business process model.
Types of business process modeling
Modeling business processes can have a different focus. It depends on what problems it is supposed to solve with its help. Accounting for absolutely all influences on the process can significantly complicate the model and lead to redundancy in the description of the process. To avoid this, business process modeling is divided by type. The type of simulation is selected depending on the characteristics of the process under study.
Most often, for the purposes of process improvement, the following types of modeling are used:
- Functional modeling. This type of modeling implies the description of processes in the form of interconnected, clearly structured functions. At the same time, a strict temporal sequence of functions, in the form in which it exists in real processes, is not necessary.
- Object Modeling- implies the description of processes as a set of interacting objects - i.e. production units. An object is any object that is transformed during the execution of processes.
- Simulation- with this type of business process modeling, it is meant to model the behavior of processes in various external and internal conditions with an analysis of the dynamic characteristics of processes and an analysis of the distribution of resources.
The division of modeling by type is performed to simplify the work and focus on certain characteristics of the process. In this case, for the same process can be applied different kinds modeling. This allows you to work with one type of model independently of others.
Principles of business process modeling
Business process modeling is based on a number of principles that make it possible to create adequate process models. Their observance makes it possible to describe a set of process state parameters in such a way that within one model the components are closely interconnected, while individual models remain sufficiently independent of each other.
The main principles of business process modeling are as follows:
- Decomposition principle– each process can be represented by a set of hierarchically arranged elements. In accordance with this principle, the process must be detailed into its constituent elements.
- Focus Principle– to develop a model, it is necessary to abstract from many process parameters and focus on key aspects. For each model, these aspects may be different.
- Documentation principle– the elements included in the process must be formalized and fixed in the model. Different designations must be used for different process elements. Fixing elements in the model depends on the type of modeling and the chosen methods.
- Consistency principle- all elements included in the process model must have an unambiguous interpretation and not contradict each other.
- The principle of completeness and sufficiency- before including this or that element in the model, it is necessary to evaluate its impact on the process. If the element is not essential for the execution of the process, then its inclusion in the model is not advisable, because it can only complicate the business process model.
Methods for modeling business processes
Today, there are a fairly large number of methods for modeling business processes. These methods are for different types modeling and allow you to focus on different aspects. They contain both graphical and textual tools, through which you can visualize the main components of the process and give precise definitions of the parameters and relationships of elements.
Most often in quality management business process modeling is performed using the following methods:
Flow Chart Diagram (workflow diagram) is a graphical method of representing a process in which operations, data, process equipment, etc. are depicted with special symbols. The method is used to display a logical sequence of process actions. The main advantage of the method is its flexibility. The process can be represented in many ways.
Data Flow Diagram (data flow diagram). A data flow diagram or DFD is used to show the transfer of information (data) from one operation of a process to another. DFD describes the relationship of operations through information and data. This method is the basis of the structural analysis of processes, since allows you to decompose the process into logical levels. Each process can be broken down into sub-processes at a higher level of detail. The use of DFD allows you to reflect only the flow of information, but not the flow of materials. A data flow diagram shows how information enters and exits a process, what actions change information, where information is stored in a process, and so on.
Role Activity Diagram (diagram of roles). It is used to model a process in terms of individual roles, groups of roles, and the interaction of roles in a process. A role is an abstract element of a process that performs some organizational function. The role diagram shows the degree of "responsibility" for the process and its operations, as well as the interaction of roles.
IDEF (Integrated Definition for Function Modeling) - is a whole set of methods for describing various aspects of business processes (IDEF0, IDEF1, IDEF1X, IDEF2, IDEF3, IDEF4, IDEF5). These methods are based on the SADT (Structured Analysis and Design Technique) methodology. The IDEF0 and IDEF3 methods are most often used to model business processes.
Modeling is the creation of a model, i.e., an image of an object that replaces it, in order to obtain information about this object by conducting experiments with its model.
A model in the general sense (generalized model) is a specific object created for the purpose of obtaining and (or) storing information (in the form of a mental image, description by sign means or a material system), reflecting the properties, characteristics and connections of the original object of an arbitrary nature, essential for the task , solved by the subject.
Object models are simpler systems, with a clear; structure, precisely defined relationships between the constituent parts, allowing a more detailed analysis of the properties of real objects and their behavior in different situations. Thus, modeling is a tool for analyzing complex systems and objects.
A number of mandatory requirements are put forward for models. First, the model must be adequate to the object, i.e., correspond to it as fully as possible in terms of the properties chosen for study.
Secondly, the model must be complete. This means that it should make it possible, with the help of appropriate methods and methods of studying the model, to investigate the object itself, i.e., to obtain some statements regarding its properties, operating principles, and behavior under given conditions.
The set of applied models can be classified according to the following criteria:
· method of modeling;
the nature of the system being modeled;
scale of modeling.
According to the modeling method, the following types of models are distinguished:
· analytical, when the behavior of the object of modeling is described in the form of functional dependencies and logical conditions;
· simulation, in which real processes are described by a set of algorithms implemented on a computer.
According to the nature of the modeled system, the models are divided into:
· to deterministic, in which all elements of the modeling object are constantly clearly defined;
· to stochastic, when the models include random controls.
Depending on the time factor, models are divided into static and dynamic. Static models (diagrams, graphs, data flow diagrams) allow one to describe the structure of the system being modeled, but do not provide information about its current state, which changes over time. Dynamic models make it possible to describe the development of processes occurring in the system over time. Unlike static models, dynamic models allow you to update the values of variables, the models themselves, dynamically calculate various process parameters and the results of impacts on the system.
Models can be divided into the following types:
1) Functional models - express direct relationships between endogenous and exogenous variables.
2) Models expressed using systems of equations with respect to endogenous quantities. They express balance ratios between various economic indicators (for example, a model of input-output balance).
3) Optimization type models. The main part of the model is a system of equations with respect to endogenous variables. But the goal is to find the optimal solution for some economic indicator (for example, to find such values of tax rates to ensure the maximum inflow of funds to the budget for a given period of time).
4) Simulation models - a very accurate reflection of the economic phenomenon. The simulation model allows you to answer the question: "What will happen if ...". The simulation system is a set of models simulating the course of the process under study, combined with a special system of auxiliary programs and an information base, which make it possible to quite simply and quickly implement variant calculations.
In this case, mathematical equations may contain complex, non-linear, stochastic dependencies.
On the other hand, models can be divided into controlled and predictive. Managed models answer the question: "What will happen if ...?"; “How to achieve what you want?” and contain three groups of variables: 1) variables that characterize the current state of the object; 2) control actions - variables that affect the change in this state and are amenable to purposeful choice; 3) initial data and external influences, i.e. externally set parameters and initial parameters.
In predictive models, control is not explicitly identified. They answer the questions: “What will happen if everything remains the same?”.
Further, models can be divided according to the method of measuring time into continuous and discrete. In any case, if time is present in the model, then the model is called dynamic. Most often, discrete time is used in models, because information is received discretely: reports, balance sheets and other documents are compiled periodically. But from a formal point of view, the continuous model may be easier to study. Note that in physical science there is a continuing discussion about whether the real physical time is continuous or discrete.
Usually, fairly large socio-economic models include material, financial and social sections. Material section - balances of products, production capacities, labor, natural resources. This is a section that describes the fundamental processes, this is a level that is usually poorly controlled, especially fast, because it is very inertial.
The financial section contains cash flow balances, rules for the formation and use of funds, pricing rules, etc. At this level, many controlled variables can be identified. They can be regulators. The social section contains information about people's behavior. This section introduces many uncertainties into decision-making models, since it is difficult to correctly take into account such factors as labor productivity, consumption patterns, motivation, etc.
When constructing models that use discrete time, econometric methods are often used. Among them, regression equations and their systems are popular. Lags are often used (delays in the reaction). For systems that are nonlinear in parameters, the application of the least squares method encounters difficulties.
Currently popular approaches to business reengineering processes are based on the active use of mathematical and information models.
When building any management process model, it is desirable to adhere to the following action plan:
1) Formulate the goals of studying the system;
2) Select those factors, components and variables that are the most significant for this task;
3) Take into account in one way or another extraneous factors not included in the model;
4) Evaluate the results, check the model, evaluate the completeness of the model.
The modeling process itself can be represented as a cycle, in which five stages can be distinguished:
1. Statement of the problem and its analysis - important features are highlighted
and properties of the object, the relationship of elements in the structure of the object is investigated, hypotheses are formulated, the behavior and development of the object is explained.
2. Building a model - the type of model is selected, the possibility of its application for solving the tasks is evaluated, the list of displayed parameters of the modeled object and the relationship between them is specified. For complex objects, the possibility of building several models that reflect various aspects of the object's functioning is determined.
3. Preparation of initial information - data is collected about the object (based on the study of the model). Then they are processed using the methods of probability theory, mathematical statistics and expert procedures.
4. Carrying out calculations and analyzing the results of the experiment - the reliability of the results is assessed.
5. Application of the results in practice - work with the simulated
object, taking into account its supposed properties obtained in the study of models. At the same time, it is assumed that these properties with a sufficient level of probability are actually inherent in this object. The last provision should be based on the results of the previous stage.
If the results obtained at the fifth stage are insufficient, the object itself or its environment has changed, then there is a return to the first stage and a new passage of the modeling cycle.
The use of modern computers, computing systems and networks is a powerful means of implementing simulation models and studying with their help the characteristics of the process of systems functioning. S. In some cases, depending on the complexity of the modeling object, i.e., the system S, rational use of personal computers (PC) or local area networks (LAN). In any case, the effectiveness of system research S on a software-implemented model M s first of all, it depends on the correctness of the scheme of the modeling algorithm, the perfection of the program, and only indirectly depends on specifications computer used for simulation. Of great importance in the implementation of the model on a computer is the question of the correct choice of the modeling language.
Modeling systems and programming languages. Algorithmic languages when modeling systems, they serve as an auxiliary apparatus for the development, machine implementation and analysis of the characteristics of models. Each modeling language should reflect a certain structure of concepts to describe a wide class of phenomena. Having chosen a specific language for solving the problem of modeling the process of functioning of the system, the researcher has at his disposal a carefully developed system of abstractions that provide him with a basis for formalizing the process of functioning of the system under study. processing and output of simulation results allow you to quickly and in detail analyze the possible outcomes of a simulation experiment with the M m model.
The main points that characterize the quality of modeling languages are: the convenience of describing the process of the system functioning S, ease of input of simulation input data and variation of the structure, algorithms and parameters of the model, feasibility of statistical modeling, efficiency of analysis and output of simulation results, ease of debugging and control of the simulation program, accessibility of perception and use of the language. The future of modeling languages is determined by progress in the field of creating multimedia systems for machine simulation, as well as problem-oriented information and computing systems for the purpose of modeling.
Consider the basic concepts associated with algorithmic languages and their implementation on a computer in general and modeling languages in particular.
Programming language is a set of characters recognized by the computer and denoting operations that can be implemented on the computer. At the lowest level is the main language of the machine, the program in which is written in codes that directly correspond to elementary machine actions (addition, memorization, forwarding to a given address, etc.). The next level is occupied by autocode (language ASSEMBLY) computing machine. An autocode program is made up of mnemonic symbols converted into machine codes by a special program - an assembler.
Compiler A program is called a program that takes instructions written in a high-level algorithmic language and converts them into programs in the main language of the machine or in autocode, which in the latter case are translated again using assembler.
Interpreter A program is called a program that, upon receiving instructions from the input language, immediately performs the corresponding operations, in contrast to the compiler, which converts these instructions into memorable chains of commands. Translation occurs during the entire time of the program written in the interpreter language. In contrast, compilation and assembly are single acts of translating text from the input language into the object language of the machine, after which the resulting programs are executed without repeated calls to the translator.
A program written in machine code or in a language ASSEMBLY, always reflects the specifics of a particular computer. The instructions of such a program correspond to certain machine operations and, therefore, make sense only in the computer for which they are intended, therefore such languages are called machine-oriented languages.
Most interpreter and compiler languages can be classified as procedurally oriented languages. These languages are qualitatively different from machine-oriented languages, which describe elementary computer operations and do not have a problem orientation. All procedural languages are intended for a certain class of problems, include instructions that are convenient for formulating ways to solve typical problems of this class. The corresponding algorithms are programmed in notations that are not associated with any computer.
The modeling language is a procedurally oriented language with specific features. The main modeling languages were developed as a software simulation approach to the study of the process of functioning of a certain class of systems.
Features of the use of algorithmic languages. Consider the advantages and disadvantages of using for modeling the process of functioning of systems simulation languages(JIM) and general purpose languages(NON), i.e., universal and procedurally oriented algorithmic languages. The expediency of using NIM stems from two main reasons: 1) the convenience of programming the system model, which plays a significant role in the machine implementation of modeling algorithms; 2) the conceptual orientation of the language to a class of systems, which is necessary at the stage of building a system model and choosing a general direction of research in the planned computer experiment. The practice of systems modeling shows that it is the use of NIM that largely determined the success of simulation as a method of experimental study of complex real objects.
Modeling languages allow describing the simulated systems in terms developed on the basis of the basic concepts of simulation. Before these concepts were clearly defined and formalized in JIM, there was no common ways descriptions of simulation tasks, and without them there was no connection between various developments in the field of setting simulation experiments. High-level modeling languages are a convenient means of communication between the customer and the developer of the machine model M m .
Despite these advantages of JIM, solid arguments, both technical and operational, are now being put forward against the complete abandonment of universal and procedural languages in modeling. Technical objections to the use of JIM: questions of the effectiveness of working programs, the possibility of debugging them, etc. As operational shortcomings, the lack of documentation on existing JIM, the purely individual nature of the corresponding translators, which complicates their implementation on various computers, and the difficulty of correcting errors are mentioned. The decrease in the efficiency of NIM is manifested when modeling problems that are more diverse than those for which a specific modeling language is designed. But here it should be noted that at present there is no NON that would be effective in solving problems of any class.
Serious shortcomings of JIM are manifested in the fact that, unlike the widely used LDL, translators from which are included in the software supplied by the manufacturer for all modern computers, modeling languages, with a few exceptions, were developed by individual organizations for their rather narrowly specialized needs. The corresponding translators are poorly described and adapted for use in solving problems of system modeling, therefore, despite the advantages of NIM, one has to abandon their practical application in a number of specific cases.
When creating a modeling system based on any language, it is necessary to solve the issue of synchronizing the processes in the model, since at each moment of time flowing in the system (system time), it may be necessary to process several events, i.e., a pseudo-parallel organization of the simulated processes in the machine model is required M m . This is the main task of the simulation monitor, which performs the following functions: process control (coordination of system and machine time) and resource management (selection and distribution of limited simulation system tools in the model).
Approaches to the development of modeling languages. To date, there have been two different approaches to the development of modeling languages: continuous and discrete - reflecting the main features of the systems studied by the modeling method. Therefore, NIM are divided into two independent groups, which correspond to two types of imitation that developed independently of each other: to simulate continuous and discrete processes.
For modeling continuous processes, not only AVM, but also computers, the latter, with appropriate programming, imitate various continuous processes. At the same time, computers are more reliable in operation and allow obtaining high accuracy of results, which led to the development of modeling languages that display the model in the form of blocks of such types that play the role of standard blocks. AVM(amplifiers, integrators, function generators, etc.). The given scheme of the modeling algorithm is transformed into a system of jointly considered differential equations. Modeling in this case is essentially reduced to finding numerical solutions to these equations using some standard step-by-step method.
An example of a language for modeling continuous systems on a computer by representing the modeled system in the form of equations in finite differences is the language DYNAMO, for which the equations establish relationships between the values of the functions at the instants of time t and t+dt and between the values of their derivatives at time t+dt/2. And in this case, the simulation, in essence, is a step-by-step solution of a given system of differential equations .
Universal computer- a device of a discrete type, and therefore should provide a discrete approximation of the process of functioning of the system under study S. Continuous changes in the process of functioning of a real system are displayed in a discrete model M m, implemented on a computer, by a certain sequence of discrete events, and such models are called discrete event models. Individual events reflected in a discrete model can be determined with a high degree of approximation to reality, which ensures the adequacy of such discrete models to real processes occurring in systems S.
Architecture of modeling languages. JIM architecture, i.e., the concept of the interrelationships of the elements of the language as complex system, and technology of transition from the system S to her machine model M s can be represented as follows: 1) modeling objects (systems S) are described (displayed in the language) using some language attributes; 2) attributes interact with processes that are adequate to the real phenomena in the simulated system S; 3) processes require specific conditions that determine the logical basis and sequence of interaction of these processes in time; 4) conditions affect the events that take place inside the simulation object (system 5) and when interacting with external environment E; 5) events change the state of the system model M in space and in time.
A typical diagram of the NIM architecture and the technology of its use in system modeling is shown in fig. 5.1.
In most cases, machine models are used to study the characteristics and behavior of the system. S over a certain period of time, therefore one of the most important tasks when creating a system model and choosing a programming language for the model, two functions are implemented: 1) adjusting the time coordinate of the system state (“advancing” time, organizing “clocks”); 2) ensuring the consistency of various blocks and events in the system (synchronization in time, coordination with other blocks).
Thus, the functioning of the Mm model should proceed in artificial (not in real and not in computer) time, ensuring the occurrence of events in the order required by the logic of the system under study and with appropriate time intervals between them. At the same time, it should be taken into account that the elements of a real system S function simultaneously (in parallel), and the components of the machine model M m act sequentially, as they are implemented using a sequential computer. Since events can occur simultaneously in different parts of the modeling object, then in order to maintain the adequacy of cause-and-effect temporal relationships, it is necessary to create a "mechanism" for setting time in JIM to synchronize the actions of the elements of the system model.
Setting the time in the machine model. As already noted in Chap. 3, there are two main approaches to setting time: using constant and variable time intervals, which correspond to two principles for the implementation of modeling algorithms, i.e., "principle D t" and "principle d z".
Consider the appropriate methods of time management in the system model M(S) on the example shown in Fig. 5.2, where the sequence of events in the system is plotted along the real-time axis ( s i) in time, and the events s 4 and s 5 occur simultaneously (Fig. 5.2, a). Driven by events s i model states change z i at the time t zi, and such a change occurs abruptly dz.
In a model built according to the "principle D t"(Fig. 5.2, b), the system time moments will sequentially take the values:
t " 1 = D t, t " 2 = 2D t, t " 3 = 3D t, t " 4 = 4D t, t " 5 = 5D t.
These moments of system time t " j(D t) are in no way related to the moments of occurrence of events s i, which are simulated in the system model. In this case, the system time receives a constant increment, which is selected in the time specified before the start of the simulation experiment.
In a model built according to the "principle dz"(Fig. 5.2, in), the time change occurs at the moment of system state change, and the sequence of system time moments has the form t "" 1 = t z 1 , t "" 2 = t z 2 , t "" 3 = t z 3 , t "" 4 = t z 4 , t "" 5 = t z 5 , i.e. points of system time t "" k (dz), are directly related to the moments of occurrence of events in the system s i .
Each of these methods has its own advantages in terms of adequate reflection of real events in the system. S and the cost of machine resources for modeling.
When using the "principle d z" events are processed sequentially and the time is shifted each time forward until the start of the next event. In a model built according to the "principle D t", event processing occurs by groups, batches or sets of events. In this case, the choice of D t has a significant impact on the course of the process and the results of the simulation, and if D t is set incorrectly, the results may turn out to be unreliable, since all events appear at the point corresponding to the upper limit of each simulation interval. When applying the "principle d z" Simultaneous processing of events in the model takes place only when these events appear simultaneously in the real system. This avoids the need to artificially introduce the ranking of events when they are processed at the end of the interval. At.
When modeling according to the "principle D t" a good approximation can be achieved: for this D t must be small so that two non-simultaneous events do not fall into the same time interval. But a decrease in D t leads to an increase in the cost of computer time for modeling, since a significant part is spent on adjusting the "clock" and tracking events, which may not occur in most intervals. In this case, even with a strong "compression" D t two non-simultaneous events can fall into the same time interval D t, which creates a false impression of their simultaneity.
To choose the principle of constructing a machine model M m and accordingly, JIM needs to know: the purpose and purpose of the model; the required accuracy of the simulation results; the cost of computer time when using one or another principle; the required amount of machine memory to implement a model built according to the D principle t and d z; the complexity of programming the model and its debugging.
Requirements for simulation languages. Thus, when developing system models, a number of specific difficulties arise, therefore, a set of such software tools and concepts that are not found in conventional NON should be provided in NIM.
Combination. Parallel flowing in real systems S processes are represented by a sequentially operating computer. Modeling languages get around this difficulty by introducing the concept of system time, which is used to represent time-ordered events.
The size. Most of the simulated systems have a complex structure and behavior algorithms, and their models are large in volume. Therefore, dynamic memory allocation is used when the components of the system model M m appear in random access memory computer or leave it, depending on the current state. An important aspect the realizability of the M m model on a computer in this case is the block nature of its design, i.e., the possibility of splitting the model into blocks, subblocks, etc.
Changes. Dynamic systems are associated with movement and are characterized by the development of the process, as a result of which the spatial configuration of these systems undergoes changes over time. Therefore, in all RIMs, they provide for the processing of lists that reflect changes in the states of the process of functioning of the simulated system S.
Interconnectedness. Conditions required for various events to occur in the model M m system operation process S, can be very difficult due to the presence a large number mutual relations between the components of the model. To resolve the difficulties associated with this issue, most JIMs include the corresponding logical possibilities and concepts of set theory.
Stochasticity. To simulate random events and processes, special programs are used to generate sequences of pseudo-random numbers, quasi-uniformly distributed over a given interval, on the basis of which it is possible to obtain stochastic effects on the M m model, imitated by random variables with the corresponding distribution law.
Analysis. To obtain a clear and practical answer to the questions solved by the method of machine simulation, it is necessary to obtain statistical characteristics of the process of functioning of the system model M(S). Therefore, modeling languages provide methods for statistical processing and analysis of modeling results.
The listed requirements in the study and design various systems S correspond to such well-known discrete event modeling languages as SIMULA, SIMSCRIPT, GPSS, S.O.L. CSL and etc.
To the means software
relate:
management process modeling tools;
typical control tasks;
methods of mathematical programming, mathematical statistics, queuing theory, etc.
Part software
includes systemic and applied software, as well as technical documentation
System software
includes Operating Systems for used hardware platforms, various operating shells that increase the level of the user interface, programming systems, programs for working in the network, system tests, programs for administering networks, databases.
Application software
can be generic or specialized.
task-oriented. It can be customized for a specific use case. DBMS, word processors, spreadsheets, text and speech recognition programs, report generators for database systems, etc. are used as such tools.
Specialized software
created for a specific information system or for a class of systems with a narrow purpose.
Typical application software
can be general-purpose or domain-specific, as well as hardware-platform-specific or mobile.
The technical documentation for the software should contain a description of the tasks, an economic and mathematical model of the task, a list of software modules of the program algorithm, a list of used symbols, test cases.
Information Support
The purpose of the information support subsystem is the modern formation and issuance of reliable information for adoption. management decisions.
Information Support
The totality of a unified system for classifying and coding information, unified documentation systems, schemes for information flows circulating in an organization, as well as a methodology for building databases.
To linguistic support of IP
includes natural and artificial languages, as well as the means of their linguistic support: dictionaries of the vocabulary of natural languages, thesauri (special dictionaries of the basic concepts of the language, denoted by individual words or phrases, with certain semantic relationships between them) of the subject area, translation dictionaries, etc.
Organizational support- a set of methods and tools that regulate the interaction of employees with technical means and among themselves in the process of developing and operating the information system.
Organizational support implements the following functions:
analysis of the existing management system of the organization, where IS will be used, and identification of tasks to be automated;
preparation of tasks for solving on a computer, including technical task for the design of IS and the feasibility study of its effectiveness;
development of management decisions on the composition and structure of the organization, methodology for solving problems aimed at improving the efficiency of the management system.
Organizational support. EIS includes its own control apparatus, which ensures the functioning and development of all subsystems. Its main functions are to develop.
For beginners: breeding a broiler at home Boiled water for broilers
Only lovers will survive
Features of advertising aimed at children
retouching old photos in photoshop retouching old photos
What is an NPO: decoding, definition of goals, types of activities Does a non-profit organization have the right