Complex decision operation optimization method based on model analog simulation
The method addresses instability in operations optimization by using model simulation to balance multiple objectives and adapt to changing environments, enhancing decision-making accuracy and efficiency through precise modeling and intelligent algorithms.
Patent Information
- Application Number
- CN202411219276.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-02
- Publication Date
- 2024-12-20
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Existing operations optimization algorithms struggle to achieve optimal balance between multiple objectives in dynamic and uncertain environments, leading to instability and increased maintenance costs due to incompatibility with changing problem structures and weights, affecting user experience and operational efficiency.
A method involving model simulation for complex decision-making, including data collection, random simulation, multi-objective optimization, sensitivity analysis, intelligent algorithms, and dynamic programming to establish precise models, identify key factors, and adapt to changing scenarios.
Enhances decision-making accuracy, adaptability, and stability by balancing multiple objectives, identifying key factors, and optimizing resource allocation, thereby improving efficiency and reducing risks in dynamic environments.
Smart Images
Figure CN119167765A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of decision-making and operational optimization methods, and more specifically, to a complex decision-making and operational optimization method based on model simulation. Background Art
[0002] Operations research is a discipline that uses mathematical methods to study the optimization paths and solutions of various systems and provide scientific basis for decision makers. It is widely used in many industries and fields. With the development of science and technology, the advancement of computing power and algorithms and the digitization of traditional industries have brought about massive data, gradually forming an intelligent manufacturing technology system with "data + computing power + algorithm" as the core. The operations optimization algorithm is the core of this technology system, and the value of data needs to be reflected through it.
[0003] However, most problems in actual production and decision-making systems are multi-objective combinatorial optimization problems, that is, non-deterministic polynomial problems or non-convex optimization problems. The two commonly used objectives are efficiency and effectiveness, but it is usually difficult to achieve the optimal goal at the same time, so the weighted method is often used to convert it into a single objective.
[0004] In actual application scenarios, when the problem structure and the weights of the solution objectives change, the operations optimization algorithms originally designed based on known problems, multi-objective weights and data characteristics, including the algorithm's hyperparameters and solution strategies, may not be compatible with the current scenario and cannot perform stably and obtain stable optimization results. To deal with this situation, users usually need to notify developers to improve and optimize the algorithm for the new scenario after discovering that the algorithm is not compatible, and then redeploy, test and put it online. This not only seriously affects the user experience, but also affects normal operations when it involves core production systems, significantly increases the cost of algorithm development and maintenance, and brings many problems to users and service providers. Summary of the invention
[0005] In order to overcome the above-mentioned defects of the prior art, the present invention provides a method for complex decision-making operations optimization based on model simulation to solve the problems raised in the above-mentioned background technology.
[0006] To achieve the above object, the present invention provides the following technical solution: a method for complex decision-making operations optimization based on model simulation, specifically comprising the following steps: Step 1: Establish an accurate mathematical model, collect data and information about the decision problem, collect information about the background, goals and constraints of the decision, define key variables that can describe the decision problem, set constraints that take into account the actual situation, determine the decision goal, and select a planning method to build the model based on the nature of the problem; Step 2: Use random simulation to determine the random variables of market demand fluctuations and equipment failure probability in decision-making, set distribution methods for random variables, design simulation experiments, determine the number of simulations and parameter settings, conduct simulation runs, use computer programs to generate random numbers, simulate decision-making results under different circumstances, analyze simulation results, calculate statistical indicators, mean, standard deviation and probability distribution, and evaluate the risks and benefits of decision-making plans; Step 3: Multi-objective optimization: clarify cost, quality and time, standardize objectives of different dimensions and magnitudes, select multi-objective optimization methods, solve optimization problems, obtain a set of non-inferior solutions for decision analysis, and the decision maker selects the final decision plan from the non-inferior solutions based on actual conditions and preferences; Step 4: Sensitivity analysis: select the cost coefficient and demand growth rate, set the parameter value range, perform sensitivity calculation, change the parameter value, recalculate the decision result, analyze the sensitivity result, observe the sensitivity of the decision result to the parameter change, and determine the key factors; Step 5: Based on the optimization of intelligent algorithms, according to the decision problem data and information collected in step 1, select an algorithm, convert the solution to the problem into a coding form that the algorithm can handle, set the algorithm parameters, run the algorithm, and let the algorithm search and optimize in the solution space to obtain the optimal or approximately optimal solution; Step 6: Combine the simulation model and optimization algorithm to establish a simulation model that describes the system operation mechanism and logic. Use software tools to build and run the simulation model to generate system performance data under different decision-making schemes. Perform pre-processing such as cleaning, sorting and analyzing the simulation data. Use the data for optimization algorithms. Use the simulation data as input to train and optimize the decision model. Through repeated iterative optimization, adjust the simulation model and parameters according to the optimization results until satisfactory results are obtained. Step 7: Dynamic programming method, divide the decision-making process into different stages according to time or logical order, define the state variables of each stage to describe the characteristics of the problem, establish the connection between the states of adjacent stages and the recursive relationship of decision-making, calculate from the last stage forward, and gradually solve the optimal decision of each stage according to the recursive relationship to obtain the overall optimal decision. The optimal decision of each stage is integrated to form the optimal strategy for the entire problem; Step 8: Scenario analysis, identify the key factors that may affect the decision, such as policy changes, market trends, and technological development, imagine different combinations and developments of key factors, construct multiple possible future scenarios, formulate decision plans for each scenario, use models to simulate and analyze the decision effects under different scenarios, evaluate and compare the plans under each scenario, comprehensively consider their performance, evaluate their adaptability and robustness, and select the final decision plan that performs well in multiple scenarios.
[0007] Preferably, the key variables of the decision problem in step one are specifically quantity, price and time, and the constraints in step one are specifically resource availability, production capacity and demand constraints.
[0008] Preferably, the decision-making objectives in step one are specifically minimizing costs, maximizing profits and improving service levels, and the planning method in step one is one or more of linear programming, integer programming and nonlinear programming.
[0009] Preferably, the distribution method in step 2 is specifically probability distribution, normal distribution and Poisson distribution, and the number of simulation experiments and parameter settings in step 2 need to be balanced between accuracy and computational cost to ensure that valuable information can be provided for decision-making.
[0010] Preferably, the standardization process in step 3 is specifically Z-score standardization, which is used for unified comparison and weighing.
[0011] Preferably, the multi-objective optimization method in step three is specifically a weighted method, an ε-constraint method or objective programming.
[0012] Preferably, the sensitivity calculation in step 4 is used to help determine which parameters have a significant impact on the decision-making results, thereby identifying the key factors in the decision-making, and simultaneously performing risk assessment, solution optimization, decision support and resource allocation.
[0013] Preferably, the algorithms in step five are specifically genetic algorithms, particle swarm optimization algorithms and ant colony algorithms. Genetic algorithms are used to solve complex optimization problems. Particle swarm optimization algorithms are used to find optimal solutions by simulating the movement of individuals in the solution space and information sharing. Ant colony algorithms are used for complex problems of path selection and combinatorial optimization.
[0014] Preferably, the algorithm parameters in step 5 are specifically population size, number of iterations and crossover probability.
[0015] Technical effects and advantages of the present invention: 1. By establishing precise mathematical models and collecting comprehensive data and information, we can accurately describe decision-making problems, reduce the blindness and subjectivity of decision-making, and thus improve the accuracy of decision-making. In the production planning of enterprises, we can more accurately determine the production quantity and time schedule to avoid overproduction or shortages. By using random simulation to consider random variables such as market demand fluctuations and equipment failure probability, we can better cope with the uncertainty in reality, enhance the adaptability and stability of decision-making, formulate corresponding emergency strategies in advance, reduce risks, and use multi-objective optimization and standardized processing to find a balance between multiple conflicting goals such as cost, quality, and time to meet the needs of different aspects; 2. Through sensitivity analysis, it helps to identify the key factors that have a significant impact on the decision-making results, which is helpful for the rational allocation and focus of resources, and to clarify which factors have the greatest impact on the return on investment, so as to concentrate on monitoring and management. Based on the optimization of intelligent algorithms and continuous iterative adjustments, it is possible to search for the optimal or near-optimal solution in the complex solution space and improve the efficiency of decision-making. - For logistics distribution route planning, find the shortest path, and reduce transportation costs. The dynamic programming method divides the decision-making process into stages and establishes a recursive relationship, so that the decision can be flexibly adjusted according to the situation at different stages; 3. In summary, through the mutual influence of the above-mentioned multiple effects, the method of complex decision-making operations optimization based on model simulation and emulation can improve decision-making accuracy and reduce blindness and subjectivity by establishing precise mathematical models and collecting comprehensive data and information; use random simulation to deal with uncertainty and enhance decision-making adaptability and stability; achieve multi-objective balance and meet multiple requirements such as cost, quality, and time; identify key factors through sensitivity analysis and allocate resources reasonably; use intelligent algorithms to obtain better solutions and improve decision-making efficiency; use dynamic programming to enhance decision-making flexibility; use scenario analysis to comprehensively evaluate plans and improve decision reliability. This method provides a scientific, systematic and comprehensive solution for complex decision-making problems, significantly improves the quality and effectiveness of decisions, and helps decision makers make more informed and effective decisions in complex and changing environments. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] Figure 1 It is a schematic diagram of the method flow structure of the present invention. DETAILED DESCRIPTION
[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Example
[0018] The method of complex decision-making operations optimization based on model simulation specifically includes the following steps: Step 1: Establish an accurate mathematical model, collect data and information about the decision problem, collect information about the background, goals and constraints of the decision, define key variables that can describe the decision problem, set constraints that take into account the actual situation, determine the decision goal, and select a planning method to build the model based on the nature of the problem; The key variables of the decision problem in step 1 are specifically quantity, price and time; the constraints in step 1 are specifically resource availability, production capacity and demand constraints; the decision goals in step 1 are specifically cost minimization, profit maximization and service level improvement; the planning method in step 1 is one or more of linear programming, integer programming and nonlinear planning methods; Step 2: Use random simulation to determine the random variables of market demand fluctuations and equipment failure probability in decision-making, set distribution methods for random variables, design simulation experiments, determine the number of simulations and parameter settings, conduct simulation runs, use computer programs to generate random numbers, simulate decision-making results under different circumstances, analyze simulation results, calculate statistical indicators, mean, standard deviation and probability distribution, and evaluate the risks and benefits of decision-making plans; The distribution methods in step 2 are probability distribution, normal distribution and Poisson distribution. The number of simulation experiments and parameter settings in step 2 need to balance accuracy and computational cost to ensure that valuable information can be provided for decision making. Step 3: Multi-objective optimization: clarify cost, quality and time, standardize objectives of different dimensions and magnitudes, select multi-objective optimization methods, solve optimization problems, obtain a set of non-inferior solutions for decision analysis, and the decision maker selects the final decision plan from the non-inferior solutions based on actual conditions and preferences; The standardization process in step 3 is specifically Z-score standardization, which is used for unified comparison and trade-off. The multi-objective optimization method in step 3 is specifically weighted method, ε-constraint method and objective planning. Step 4: Sensitivity analysis: select the cost coefficient and demand growth rate, set the parameter value range, perform sensitivity calculation, change the parameter value, recalculate the decision result, analyze the sensitivity result, observe the sensitivity of the decision result to the parameter change, and determine the key factors; The sensitivity calculation in step 4 is used to help determine which parameters have a significant impact on the decision-making results, thereby identifying the key factors in the decision-making, and at the same time conducting risk assessment, program optimization, decision-making assistance and resource allocation; Step 5: Based on the optimization of intelligent algorithms, according to the decision problem data and information collected in step 1, select an algorithm, convert the solution to the problem into a coding form that the algorithm can handle, set the algorithm parameters, run the algorithm, and let the algorithm search and optimize in the solution space to obtain the optimal or approximately optimal solution; The algorithms in step 5 are specifically genetic algorithm, particle swarm optimization algorithm and ant colony algorithm. Genetic algorithm is used to solve complex optimization problems. Particle swarm optimization algorithm is used to find the optimal solution by simulating the movement of individuals in the solution space and information sharing. Ant colony algorithm is used for complex problems of path selection and combinatorial optimization. The algorithm parameters in step 5 are specifically population size, number of iterations and crossover probability. Step 6: Combine the simulation model and optimization algorithm to establish a simulation model that describes the system operation mechanism and logic. Use software tools to build and run the simulation model to generate system performance data under different decision-making schemes. Perform pre-processing such as cleaning, sorting and analyzing the simulation data. Use the data for optimization algorithms. Use the simulation data as input to train and optimize the decision model. Through repeated iterative optimization, adjust the simulation model and parameters according to the optimization results until satisfactory results are obtained. Step 7: Dynamic programming method, divide the decision-making process into different stages according to time or logical order, define the state variables of each stage to describe the characteristics of the problem, establish the connection between the states of adjacent stages and the recursive relationship of decision-making, calculate from the last stage forward, and gradually solve the optimal decision of each stage according to the recursive relationship to obtain the overall optimal decision. The optimal decision of each stage is integrated to form the optimal strategy for the entire problem; Step 8: Scenario analysis, identify the key factors that may affect the decision, such as policy changes, market trends, and technological development, imagine different combinations and developments of key factors, construct multiple possible future scenarios, formulate decision plans for each scenario, use models to simulate and analyze the decision effects under different scenarios, evaluate and compare the plans under each scenario, comprehensively consider their performance, evaluate their adaptability and robustness, and select the final decision plan that performs well in multiple scenarios.
[0019] Finally, it should be noted that the drawings of the embodiments disclosed in the present invention only involve structures related to the embodiments disclosed in the present invention, and other structures can refer to common designs. In the absence of conflicts, the same embodiment and different embodiments of the present invention can be combined with each other; Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A method for complex decision-making operations optimization based on model simulation, characterized by: The specific steps include: Step 1: Establish an accurate mathematical model, collect data and information about the decision problem, collect information about the background, goals and constraints of the decision, define key variables that can describe the decision problem, set constraints that take into account the actual situation, determine the decision goal, and select a planning method to build the model based on the nature of the problem; Step 2: Use random simulation to determine the random variables of market demand fluctuations and equipment failure probability in decision-making, set distribution methods for random variables, design simulation experiments, determine the number of simulations and parameter settings, conduct simulation runs, use computer programs to generate random numbers, simulate decision-making results under different circumstances, analyze simulation results, calculate statistical indicators, mean, standard deviation and probability distribution, and evaluate the risks and benefits of decision-making plans; Step 3: Multi-objective optimization: clarify cost, quality and time, standardize objectives of different dimensions and magnitudes, select multi-objective optimization methods, solve optimization problems, obtain a set of non-inferior solutions for decision analysis, and the decision maker selects the final decision plan from the non-inferior solutions based on actual conditions and preferences; Step 4: Sensitivity analysis: select the cost coefficient and demand growth rate, set the parameter value range, perform sensitivity calculation, change the parameter value, recalculate the decision result, analyze the sensitivity result, observe the sensitivity of the decision result to the parameter change, and determine the key factors; Step 5: Based on the optimization of intelligent algorithms, according to the decision problem data and information collected in step 1, select an algorithm, convert the solution to the problem into a coding form that the algorithm can handle, set the algorithm parameters, run the algorithm, and let the algorithm search and optimize in the solution space to obtain the optimal or approximately optimal solution; Step 6: Combine the simulation model and optimization algorithm to establish a simulation model that describes the system operation mechanism and logic. Use software tools to build and run the simulation model to generate system performance data under different decision-making schemes. Perform pre-processing such as cleaning, sorting and analyzing the simulation data. Use the data for optimization algorithms. Use the simulation data as input to train and optimize the decision model. Through repeated iterative optimization, adjust the simulation model and parameters according to the optimization results until satisfactory results are obtained. Step 7: Dynamic programming method, divide the decision-making process into different stages according to time or logical order, define the state variables of each stage to describe the characteristics of the problem, establish the connection between the states of adjacent stages and the recursive relationship of decision-making, calculate from the last stage forward, and gradually solve the optimal decision of each stage according to the recursive relationship to obtain the overall optimal decision. The optimal decision of each stage is integrated to form the optimal strategy for the entire problem; Step 8: Scenario analysis, identify the key factors that may affect the decision, such as policy changes, market trends, and technological development, imagine different combinations and developments of key factors, construct multiple possible future scenarios, formulate decision plans for each scenario, use models to simulate and analyze the decision effects under different scenarios, evaluate and compare the plans under each scenario, comprehensively consider their performance, evaluate their adaptability and robustness, and select the final decision plan that performs well in multiple scenarios.
2. The method for complex decision-making operations optimization based on model simulation according to claim 1 is characterized in that: The key variables of the decision-making problem in step one are specifically quantity, price and time, and the constraints in step one are specifically resource availability, production capacity and demand restrictions.
3. The method for complex decision-making operations optimization based on model simulation according to claim 1 is characterized in that: The decision-making objectives in step one are specifically minimizing costs, maximizing profits and improving service levels, and the planning method in step one is one or more of linear programming, integer programming and nonlinear programming.
4. The method for complex decision-making operations optimization based on model simulation according to claim 1 is characterized in that: The distribution method in step 2 is specifically probability distribution, normal distribution and Poisson distribution. The number of simulation experiments and parameter settings in step 2 need to be balanced between accuracy and computational cost to ensure that valuable information can be provided for decision-making.
5. The method for complex decision-making operations optimization based on model simulation according to claim 1 is characterized in that: The standardization process in step 3 is specifically Z-score standardization, which is used for unified comparison and weighing.
6. The method for complex decision-making operations optimization based on model simulation according to claim 1 is characterized in that: The multi-objective optimization method in step three is specifically a weighted method, an ε-constraint method and objective programming.
7. The method for complex decision-making operations optimization based on model simulation according to claim 1 is characterized in that: The sensitivity calculation in step 4 is used to help determine which parameters have a significant impact on the decision-making results, thereby identifying the key factors in the decision-making, and at the same time performing risk assessment, program optimization, decision support and resource allocation.
8. The method for complex decision-making operations optimization based on model simulation according to claim 1 is characterized in that: The algorithms in step five are specifically genetic algorithm, particle swarm optimization algorithm and ant colony algorithm. The genetic algorithm is used to solve complex optimization problems. The particle swarm optimization algorithm is used to find the optimal solution by simulating the movement of individuals in the solution space and information sharing. The ant colony algorithm is used for complex problems of path selection and combinatorial optimization.
9. The method for complex decision-making operations optimization based on model simulation according to claim 1, characterized in that: The algorithm parameters in step 5 are specifically population size, number of iterations and crossover probability.
Citation Information
Cited By
Parameter optimization method for injection molding process of automobile air filter end cover
CN122616008A