Industrial solid waste multi-target collaborative management method and system based on digital twinning and hybrid optimization algorithm

Abstract

According to the industrial solid waste multi-target collaborative management method and system based on the digital twinborn and hybrid optimization algorithm, industrial solid waste data are collected in real time through an Internet of Things sensor, a dynamic digital twinborn model is constructed to simulate a physical process, and the problems of data dispersion and lag in a traditional method are solved; establishing a multi-objective optimization function by combining three major objectives of economy, environment and society, and generating a Pareto optimal strategy set by adopting an improved NSGA-II and particle swarm optimization (PSO) hybrid algorithm; the method combines historical and real-time data through a twin model, employs an LSTM algorithm to synchronously predict the type and quantity of solid wastes, the equipment operation state and the environment change, achieves the dynamic simulation and prediction of the whole process of solid waste generation, transportation and processing, remarkably improves the data integration efficiency, provides precise real-time support for decision making, and improves the real-time performance of the system. The treatment cost is reduced, the carbon emission is reduced, and the public satisfaction is improved; the strategy response time is shortened through dynamic model correction, an intelligent solution is provided for industrial solid waste management, and the function of multi-dimensional collaborative optimization of industrial solid waste management is achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of smart cities and industrial solid waste resource utilization, and specifically relates to a multi-objective collaborative management method and system for industrial solid waste based on digital twins and hybrid optimization algorithms. Background Art

[0002] In recent years, artificial intelligence technology has developed rapidly, and its application in the industrial sector has become increasingly widespread. However, correspondingly, the annual generation of industrial solid waste has exceeded 4 billion tons, but the comprehensive utilization rate is less than 60%. Therefore, the effective application of artificial intelligence technology in the management of industrial solid waste is extremely important.

[0003] However, in the field of industrial solid waste management, traditional methods mostly rely on manual experience or single-objective optimization, resulting in several significant problems. First, data collaboration is insufficient. Data on various links of industrial solid waste generation, transportation, and treatment are scattered, lacking real-time sharing and dynamic integration, which ultimately leads to a certain lag in decision-making. Secondly, there are the limitations of single-objective optimization. Existing methods mostly focus on the goal of economic efficiency (i.e., cost minimization), but ignore the coordinated optimization of indicators such as environmental and social benefits. Finally, the algorithm efficiency is low. Traditional optimization algorithms (such as genetic algorithms) are prone to falling into local optimal situations and have a slow convergence speed, making it difficult to adapt to the needs of dynamic scenarios.

[0004] To address the above issues, relevant research reports have introduced IoT technology to achieve real-time data collection and integrate solid waste management data through the interoperability of multiple modules. Although this solution aggregates data in real time, the system does not address various dimensional objectives and therefore lacks multi-dimensional modeling capabilities and collaborative optimization capabilities. Another solution, while starting from a collaborative perspective and analyzing solid waste management by constructing a random forest model, is essentially to output the result of "the direction of resource utilization" rather than solving multi-objective optimization problems. Another solution, while implementing dual-objective optimization, that is, comprehensively considering resource conversion rate and economic efficiency, still uses traditional genetic algorithms for solution, and the optimization dimension and depth of thinking in the process of optimizing objectives are limited. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a multi-objective collaborative management method and system for industrial solid waste based on digital twins and hybrid optimization algorithms, which is used for multi-dimensional collaborative optimization of industrial solid waste management.

[0006] The technical solution adopted by the present invention to solve the above technical problems is: a multi-objective collaborative management method for industrial solid waste based on digital twins and hybrid optimization algorithms, comprising the following steps: S1: Collect current data including solid waste type, solid waste quantity, equipment operating parameters and environmental data through sensors, and obtain historical data through the database; S2: Build a digital twin model at the current moment based on current and historical data; use the LSTM algorithm to dynamically predict the type and quantity of solid waste, equipment operating parameters, and environmental data at the next moment and update the digital twin model; S3: Establish a multi-objective optimization function that includes economic, environmental, and social goals. Based on the next moment's solid waste type and quantity, equipment operating parameters, and environmental data output by the digital twin model, a hybrid algorithm of NSGA-II and PSO is used to solve the multi-objective optimization function and generate a Pareto optimal solution set. S4: Dynamically correct the digital twin model through the actual data collected by the sensor and the error function.

[0007] According to the above scheme, in step S1, an Internet of Things sensor network is deployed in the industrial solid waste treatment link, and the Internet of Things sensors include weight sensors, temperature and humidity sensors, GPS positioning modules and energy consumption monitoring units.

[0008] According to the above scheme, in step S3, The economic goal is to minimize the unit solid waste treatment cost and unit distance transportation cost while maximizing resource utilization benefits. Resource utilization benefits are the recycled resource benefits obtained per unit solid waste volume, which is equal to the product of the resource utilization benefit coefficient and the resource conversion efficiency. The unit solid waste treatment cost is calculated based on the predicted solid waste type and amount; the unit distance transportation cost is calculated based on the predicted solid waste amount; and the resource conversion efficiency is calculated based on the predicted solid waste type and equipment operating parameters. The environmental goal is to minimize carbon emissions and energy consumption. Carbon emissions and energy consumption are obtained by weighting the unit solid waste treatment cost and unit distance transportation cost using different environmental goal weight coefficients and carbon emission coefficients, respectively. The carbon emission coefficient is calculated based on the predicted equipment operating parameters and environmental data. The environmental goal weight coefficient corresponds to the carbon emission coefficient. The social goal is to maximize public satisfaction, which is equal to the weighted sum of resource conversion efficiency and carbon dioxide emission reduction according to the social goal weight coefficient; carbon dioxide emission reduction is calculated in real time based on the predicted solid waste quantity and environmental data; the weight coefficient of the social goal is dynamically adjusted through the entropy weight method.

[0009] According to the above scheme, in step S3, the specific steps of using the hybrid algorithm of NSGA-II and PSO to solve the multi-objective optimization function are as follows: Initialize the population; The NSGA-II algorithm is used to generate the Pareto optimal solution set through non-dominated sorting; Calculate congestion; Determine whether the solution converges; if it converges, output the solution set; if it does not converge, execute the PSO algorithm based on the solid waste type, solid waste quantity, equipment operating parameters and environmental data predicted by the digital twin model, accelerate convergence through particle update collaboration, and jump to the step of generating the Pareto optimal solution set through non-dominated sorting using the NSGA-II algorithm.

[0010] Furthermore, in step S3, the crossover probability and mutation probability of the NSGA-II algorithm in the hybrid algorithm are adaptively adjusted with the number of iterations.

[0011] Furthermore, in step S3, the specific steps of using the PSO algorithm to accelerate convergence through particle update collaboration are as follows: Let the solution vector be the position of the particle at the current moment, and the velocity of the particle at the next iteration as the direction of change and step size of the solution; The particle's position at the current moment is subtracted from the individual cognition and social group positions respectively, and the two differences are weighted by the corresponding acceleration constant and random number respectively; then the sum is added with the particle's speed at the current moment weighted by the inertia weight to obtain the particle's speed at the next iteration.

[0012] Furthermore, in step S4, the error function is the mean of the difference between the predicted value and the actual value output by the digital twin model in each round; when the error exceeds the threshold, the learning rate of the digital twin model is adjusted using the LSTM algorithm to update the learning rate to respond to emergencies.

[0013] Industrial solid waste multi-objective collaborative management system based on digital twin and hybrid optimization algorithm, The data acquisition submodule is used to collect current data including solid waste types, solid waste quantities, equipment operating parameters and environmental data through sensors, and obtain historical data through the database; The digital twin engine submodule is used to build a digital twin model at the current moment based on current and historical data. It uses the LSTM algorithm to dynamically predict the type and quantity of solid waste, equipment operating parameters, and environmental data at the next moment and update the digital twin model. Optimize the decision-making submodule and establish a multi-objective optimization function that includes economic goals, environmental goals, and social goals. Based on the solid waste type and quantity, equipment operating parameters, and environmental data at the next moment output by the digital twin model, a hybrid algorithm of NSGA-II and PSO is used to solve the multi-objective optimization function and generate a Pareto optimal solution set. The feedback correction submodule is used to dynamically correct the parameters of the digital twin model through the actual data collected by the sensor and the error function.

[0014] Further, The network structure of the digital twin engine submodule includes an input layer, a hidden layer, and an output layer; The input layer includes features such as the types of solid waste, the amount of solid waste, equipment operating parameters, and environmental data. The hidden layer is a multi-layer LSTM, which contains a doubling number of neurons; The output layer includes branches corresponding to the predicted values ​​of solid waste types, solid waste quantities, equipment operating parameters and environmental data.

[0015] A computer memory stores a computer program that can be executed by a computer processor. The computer program executes a multi-objective collaborative management method for industrial solid waste based on digital twins and a hybrid optimization algorithm.

[0016] The beneficial effects of the present invention are: 1. The multi-objective collaborative management method and system of industrial solid waste based on digital twin and hybrid optimization algorithm of the present invention collects industrial solid waste data in real time through Internet of Things sensors, constructs a dynamic digital twin model to simulate physical processes, and solves the problems of data dispersion and lag in traditional methods; combines the three major goals of economy, environment and society to establish a multi-objective optimization function, and adopts the improved NSGA-II and particle swarm optimization (PSO) hybrid algorithm to generate the Pareto optimal strategy set; combines historical and real-time data through the twin model, and adopts the LSTM algorithm to simultaneously predict the type and quantity of solid waste, equipment operating status and environmental changes, so as to realize dynamic simulation and prediction of the entire process of solid waste generation, transportation and treatment, significantly improve data integration efficiency, provide accurate real-time support for decision-making, reduce processing costs, reduce carbon emissions, and improve public satisfaction; shortens the strategy response time through dynamic model correction, provides an intelligent solution for industrial solid waste management, and realizes the function of multi-dimensional collaborative optimization of industrial solid waste management.

[0017] 2. The present invention integrates digital twin technology and hybrid optimization algorithm to construct a dynamic multi-objective collaborative management framework, generates a Pareto optimal solution set through the improved NSGA-II and PSO hybrid algorithm, combines non-dominated sorting with the rapid convergence characteristics of particle swarms, takes into account the diversity of solution sets and global search capabilities, reduces carbon emissions while effectively balancing costs, improves public satisfaction, takes into account multi-dimensional and multi-type objectives, and overcomes the one-sidedness of traditional single-objective or dual-objective optimization.

[0018] 3. The present invention breaks through the limitations of single dimension through multi-objective collaborative optimization and achieves balanced optimization of economic, environmental and social benefits: economic goals effectively control costs from three perspectives: solid waste treatment, waste transportation and resource transformation; introduces carbon emission coefficients and environmental goal weight coefficients into environmental goals, combines transportation routes, equipment parameters, environmental data and other contents for real-time optimization, and effectively reduces energy consumption and environmental load; dynamically adjusts the social goal weight coefficient based on the key parameters in the digital twin model to ensure the balanced optimization between improving resource conversion efficiency and reducing carbon dioxide emissions; fills the technical gap in intelligent decision-making for industrial solid waste, effectively controls costs, promotes the transformation of industrial solid waste management towards green and low-carbon, and meets the common needs of people.

[0019] 4. To address the problem that traditional optimization algorithms are prone to falling into local optimality and have slow convergence, this invention adopts a hybrid algorithm to adaptively adjust the crossover and mutation probabilities, combined with a dynamic correction mechanism for model parameters, to optimize dynamic response and algorithm efficiency. At the same time, the digital twin model uses the LSTM algorithm to update the learning rate in real time, quickly responding to sudden changes in the solid waste treatment process (such as a sudden increase in solid waste volume), shortening the strategy adjustment time, and enhancing the system's adaptability in dynamic scenarios.

[0020] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0022] Figure 1 is a flow chart of an embodiment of the present invention.

[0023] Figure 2 2 is a closed-loop system architecture diagram of an embodiment of the present invention.

[0024] Figure 3 This is a workflow diagram of the digital twin engine of an embodiment of the present invention.

[0025] Figure 4 It is a flow chart of the hybrid optimization algorithm of the optimization decision platform according to an embodiment of the present invention.

[0026] Figure 5 3. It is a two-dimensional Pareto front comparison diagram of the embodiment of the present invention and comparative example 1.

[0027] Figure 6 3D Pareto front diagram of an embodiment of the present invention. DETAILED DESCRIPTION

[0028] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0029] Example 1 See also Figure 1 The specific steps of the multi-objective collaborative management method of industrial solid waste based on digital twins and hybrid optimization algorithm are as follows: S1: Collect current data including solid waste type, solid waste quantity, equipment operating parameters and environmental data through sensors, and obtain historical data through the database; S2: Build a digital twin model at the current moment based on current and historical data; use the LSTM algorithm to dynamically predict the type and quantity of solid waste, equipment operating parameters, and environmental data at the next moment and update the digital twin model; S3: Establish a multi-objective optimization function that includes economic, environmental, and social goals. Based on the next moment's solid waste type and quantity, equipment operating parameters, and environmental data output by the digital twin model, a hybrid algorithm of NSGA-II and PSO is used to solve the multi-objective optimization function and generate a Pareto optimal solution set. S4: Dynamically correct the digital twin model through the actual data collected by the sensor and the error function.

[0030] According to the above scheme, in step S1, an Internet of Things sensor network is deployed in the industrial solid waste treatment link, and the Internet of Things sensors include weight sensors, temperature and humidity sensors, GPS positioning modules and energy consumption monitoring units.

[0031] Furthermore, in step S3, The economic goal is to minimize the unit solid waste treatment cost and unit distance transportation cost while maximizing resource utilization benefits. Resource utilization benefits are the recycled resource benefits obtained per unit solid waste volume, which is equal to the product of the resource utilization benefit coefficient and the resource conversion efficiency. The unit solid waste treatment cost is calculated based on the predicted solid waste type and amount; the unit distance transportation cost is calculated based on the predicted solid waste amount; and the resource conversion efficiency is calculated based on the predicted solid waste type and equipment operating parameters. The environmental goal is to minimize carbon emissions and energy consumption. Carbon emissions and energy consumption are obtained by weighting the unit solid waste treatment cost and unit distance transportation cost using different environmental goal weight coefficients and carbon emission coefficients, respectively. The carbon emission coefficient is calculated based on the predicted equipment operating parameters and environmental data. The environmental goal weight coefficient corresponds to the carbon emission coefficient. The social goal is to maximize public satisfaction, which is equal to the weighted sum of resource conversion efficiency and carbon dioxide emission reduction according to the social goal weight coefficient; carbon dioxide emission reduction is calculated in real time based on the predicted solid waste quantity and environmental data; the weight coefficient of the social goal is dynamically adjusted through the entropy weight method.

[0032] In step S3, the specific steps of using the hybrid algorithm of NSGA-II and PSO to solve the multi-objective optimization function are as follows: Initialize the population; The NSGA-II algorithm is used to generate the Pareto optimal solution set through non-dominated sorting; Calculate congestion; Determine whether the solution converges; if it converges, output the solution set; if it does not converge, execute the PSO algorithm based on the solid waste type, solid waste quantity, equipment operating parameters and environmental data predicted by the digital twin model, accelerate convergence through particle update collaboration, and jump to the step of generating the Pareto optimal solution set through non-dominated sorting using the NSGA-II algorithm.

[0033] Furthermore, in step S3, the crossover probability and mutation probability of the NSGA-II algorithm in the hybrid algorithm are adaptively adjusted with the number of iterations.

[0034] Furthermore, in step S3, the specific steps of using the PSO algorithm to accelerate convergence through particle update collaboration are as follows: Let the solution vector be the position of the particle at the current moment, and the velocity of the particle at the next iteration as the direction of change and step size of the solution; The particle's position at the current moment is subtracted from the individual cognition and social group positions respectively, and the two differences are weighted by the corresponding acceleration constant and random number respectively; then the sum is added with the particle's speed at the current moment weighted by the inertia weight to obtain the particle's speed at the next iteration.

[0035] In step S4, the error function is the mean of the difference between the predicted value and the actual value output by the digital twin model in each round; when the error exceeds the threshold, the learning rate of the digital twin model updated by the LSTM algorithm is adjusted to respond to emergencies.

[0036] This embodiment uses IoT sensors to collect industrial solid waste data in real time, constructs a dynamic digital twin model to simulate physical processes, and solves the problems of data dispersion and lag in traditional methods. It combines economic, environmental, and social goals to establish a multi-objective optimization function, and uses an improved NSGA-II and particle swarm optimization (PSO) hybrid algorithm to generate a Pareto optimal strategy set. By combining historical and real-time data through the twin model, dynamic simulation and prediction of the entire process of solid waste generation, transportation, and treatment are achieved, significantly improving data integration efficiency, providing accurate real-time support for decision-making, reducing processing costs, reducing carbon emissions, and improving public satisfaction. Dynamic model correction shortens strategy response time, provides an intelligent solution for industrial solid waste management, and realizes the function of multi-dimensional collaborative optimization of industrial solid waste management.

[0037] Example 2 The steps of this embodiment are the same as those of embodiment 1, except that each step is applied to a specific example. Specifically, the following steps are included: S1: Deploy IoT sensor networks in industrial solid waste treatment to collect the types of solid waste at the current moment. S t , amount of solid waste Q t , Equipment operating parameters P t and environmental data E t ; Get t through the database - Historical data at 1 moment H t-1 ; Types of solid waste S The specific types of solid waste are coded; the amount of solid waste Q is the actual weight of the solid waste, in tons; equipment operating parameters P Including crusher speed, incinerator temperature, etc.; environmental data E Contains temperature, humidity, etc. IoT sensors include weight sensors, temperature and humidity sensors, GPS positioning modules and energy consumption monitoring units.

[0038] S2: Build a dynamic digital twin model based on current data and historical data. The model expression is: M t = f ( S t , Q t , P t , E t , Ht-1 ) in H t-1 For historical data, M t A digital twin model of the current moment; The key parameters of the digital twin model are dynamically predicted by the LSTM algorithm to predict the type of solid waste at the next moment. S t+1 , amount of solid waste Q t+1 , Equipment operating parameters P t+1 and environmental data E t+1 The prediction formula is:

[0039] in, S t is the type of solid waste at the current moment, Q t is the amount of solid waste at the current moment, P t The device operating parameters at the current moment, E t is the environmental data at the current moment. The network structure includes input layer, hidden layer and output layer; the input layer contains n features, namely S, Q, P, E The hidden layer is a 3-layer LSTM with 128, 64, and 32 neurons respectively; the output layer consists of 4 branches, corresponding to S, Q, P, E The predicted value of .

[0040] S3: Establish a multi-objective optimization function including economic goals, environmental goals and social goals; use the hybrid algorithm of NSGA-II algorithm and PSO algorithm to solve the multi-objective optimization function; the specific steps are: S31: Establish a multi-objective optimization function, including the following sub-objectives: Economic goal: Minimize unit solid waste treatment cost C 处理 and transportation cost per unit distance C 运输 , maximizing resource benefits k·η 转化 , the expression is: min F 1 = C 处理 ( S t+1 , Q t+1 ) + C 运输 ( Qt+1 ) −k·η 转化 ( S t+1 , P t+1 ) in, C 处理 is the unit solid waste treatment cost (yuan / ton), depending on the type of solid waste S t+1 and the amount of solid waste Q t+1 ; C 运输 is the unit distance transportation cost (yuan / km), and the amount of solid waste Q t+1 Related; k is the resource utilization benefit coefficient (10,000 yuan / unit efficiency), η 转化 is the resource conversion efficiency (%), and the calculation formula is η 转化 = Recycled resources / total solid waste × 100%, this parameter η 转化 Types of solid waste S t+1 and equipment operating parameters P t+1 related; Environmental goal: Minimize carbon emissions and energy consumption, expressed as: min F 2 =α·m·C 运输 +β·n·C 处理 in, α 、 β are all environmental target weight coefficients, m and n All are carbon emission coefficients; carbon emission coefficients m and n Equipment operating parameters output by the digital twin model P t+1 and environmental data E t+1 Related (generally speaking, the environmental target weight coefficient corresponds to the carbon emission coefficient); Social goal: maximize public satisfaction, including resource conversion efficiency and carbon dioxide emission reduction, expressed as: max F 3 =w 1 ·η 转化 ( S t+1 ,P t+1 ) +w 2. ΔCO 2 ( Q t+1 , E t+1 ) in, η 转化 Represents the resource conversion efficiency mentioned above and the type of solid waste S t+1 and equipment operating parameters P t+1 Related; Carbon dioxide emission reduction (tons) ΔCO 2According to the amount of solid waste Q t+1 and environmental data E t+1 Real-time calculation; social goal weight coefficient w 1 and w 2 Dynamically adjusted by entropy weight method, the calculation formula is: ,

[0041] in, ej is the entropy value, pij is the standardized index value.

[0042] S32: Using a hybrid strategy of NSGA-II algorithm and PSO algorithm to solve multi-objective optimization problems; see Figure 4 , the specific steps are: Initialize the population; The NSGA-II algorithm is used to generate the Pareto optimal solution set through non-dominated sorting; the crossover probability of NSGA-II in the hybrid algorithm P c and mutation probability P m Adaptively adjust with the number of iterations, the formula is: ,

[0043] in, t is the current iteration number, T is the total number of iterations; Calculate congestion; Determine whether the solution has converged; if so, output the solution set; if not, execute the PSO algorithm based on the solid waste type, solid waste quantity, equipment operating parameters, and environmental data predicted by the digital twin model, accelerate convergence through particle update collaboration, and jump to the step of generating the Pareto optimal solution set through non-dominated sorting using the NSGA-II algorithm; The particle update formula of PSO is:

[0044] in, For particles i The speed at the k+1th iteration (the direction of change of the solution and the step size); w is the inertia weight, which is used to balance the global and local search capabilities (usually decreases with iteration); c 1 and c 2 is the acceleration constant, which is used to control individual cognition ( ) and social groups ( ) r 1 and r 2 is a random number in the range [0,1], used to increase the randomness of the search; is the historical optimal position of the particle; is the current local optimal position of the particle; For particles i The position at the kth iteration (solution vector); S4: Dynamically correct the digital twin model parameters through the error function.

[0045] Assume the error function is:

[0046] in, Y 实际 is the actual observed value (such as the actual amount of solid waste, etc.), Y 预测 For the model prediction value (such as the output of LSTM Q t+1 ).when ε When the threshold is exceeded, the adaptive learning rate of the LSTM algorithm is adjusted. λ (0.6≤ λ ≤0.9) Update the model. The digital twin model at the next moment after the update is: M t+1 =M t +λ ·( Y 实际 −Y 预测 ).

[0047] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0048] Example 3 This example illustrates the solid waste management process at an industrial park. The scenario involves an annual solid waste processing capacity of 100,000 tons. The equipment includes crushers, incinerators, and a transport fleet. Sensor deployment includes a weight sensor (HBM S9M, ±0.5kg accuracy, installed at the solid waste collection point), a temperature and humidity sensor (Sensirion SHT45, ±1.5% error, deployed at the incinerator entrance), and a GPS module (u-blox NEO-M8N, for real-time tracking of transport vehicle locations).

[0049] The digital twin model was built based on MATLAB Simulink, and the overall prediction error of LSTM was ≤3%.

[0050] Optimize and solve the above model: For economic goal optimization, it predicts the type of solid waste based on the LSTM algorithm S t+1 , amount of solid waste Q t+1 etc. to dynamically adjust the transport route (GPS data), where processing costs C 处理 Depending on the type of solid waste S t+1 and the amount of solid waste Q t+1 ; Transportation costs C 运输 and the amount of solid waste Q t+1 Related; Resource conversion efficiency η 转化 And solid waste types S t+1 and equipment operating parameters P t+1 related; For environmental target optimization, its carbon emission calculation is real-time and consistent with equipment operating parameters. P t+1 (crusher speed, incinerator temperature, etc.) and environmental data E t+1 (temperature, humidity, etc.); For social goal optimization, its resource conversion efficiency η 转化 By solid waste type S t+1 and equipment operating parameters P t+1 Matching optimal process; CO2 emission reduction ΔCO 2According to the amount of solid waste Q t+1 and environmental data E t+1 Real-time calculation.

[0051] The final result is as follows: From an economic perspective, the dynamic, real-time updating of key parameters reduces the frequency and likelihood of empty transports, reducing costs from 1.2 million yuan / month to 980,000 yuan / month, an 18% reduction in monthly unit costs. From an environmental perspective, the coordinated presentation of equipment operating parameters and environmental data has improved the efficiency of solid waste transportation and treatment, reducing carbon emissions from 450 tons to 306 tons, a 32% reduction in total carbon emissions. As for social indicators, due to the combined effect of the above four key parameters, the process can balance the relationship between resource conversion efficiency and carbon dioxide emission reduction, and the final public satisfaction score is increased to a high score of 85.6.

[0052] Comparative Example 1 The difference between this comparative example and Example 3 is that: (1) Difference in application scenarios: This comparison only optimizes economic objectives; (2) Difference in optimization methods: This comparative example does not use a digital twin model to predict key parameters and uses a traditional genetic algorithm for optimization.

[0053] The final result is: From an economic perspective, the lack of real-time updates of key parameters may lead to an increase in empty loads or inefficient transportation. Although the final unit cost dropped from 1.2 million yuan / month to 1.08 million yuan / month, a 10% reduction, it is still less efficient than Example 3. In terms of environmental benefits, due to the neglect of the impact of equipment operating parameters and environmental data, there was a lack of dynamic correction of the carbon emission coefficient, resulting in the final carbon emissions increasing to 495 tons, that is, the total carbon emissions increased by 10%; In terms of social benefits, due to the neglect of the coupling effect of the key parameters mentioned above, the final public satisfaction score dropped to 65 points.

[0054] See also Figure 5 This is a comparison diagram of the two-dimensional Pareto front of economic-environmental target optimization of Example 3 and Comparative Example 1. The solution set of Example 3 is evenly distributed, while the solution set of Comparative Example 1 is concentrated in the high-cost-high carbon emission area.

[0055] Example 4 The application scenario of this embodiment is the response to a sudden increase in solid waste in a factory. Specifically, the solid waste volume of a chemical plant suddenly increased by 50%. The results of its dynamic response are as follows: (1) Model update: learning rate λ =0.8, error threshold ε ≤5%; (2) Strategy adjustment: Using key parameters in the digital twin model to plan transportation routes in real time (resulting in a 15-kilometer reduction in detours) and dynamically allocate equipment loads (resulting in an increase in incinerator utilization from 60% to 80%). The final result is: In terms of computational cost, due to the dynamic adjustment of key parameters in the digital twin model, the processing time of this embodiment is shortened to 2.5 hours. Compared with the traditional method (processing time 8 hours), the processing time is shortened by about 70%, and the increase in carbon emissions is only 5%. The public satisfaction score remains at 83 points.

[0056] See also Figure 6 The three-dimensional Pareto front diagram for optimizing the economic, environmental and social objectives of Example 4 shows that the solution set of Example 4 is located in the low-cost-low-carbon-emission-high-satisfaction area.

[0057] The above comparison results show that in order to address the problem that traditional optimization algorithms are prone to falling into local optimality and have slow convergence speed, this embodiment adopts a hybrid algorithm to adaptively adjust the crossover and mutation probabilities, combined with a dynamic correction mechanism for model parameters, to optimize dynamic response and algorithm efficiency; at the same time, the digital twin model updates the learning rate in real time through error feedback, quickly responding to sudden changes in the solid waste treatment process (such as a sudden increase in solid waste volume), shortening the strategy adjustment time, and enhancing the adaptability of the system in dynamic scenarios.

[0058] Example 5 This embodiment is used to implement the principles of the above method embodiments to build a multi-objective collaborative management system for industrial solid waste based on digital twins and hybrid optimization algorithms. Figure 2 This is a four-layer closed-loop system architecture diagram, including data acquisition module, digital twin engine, optimization decision platform and feedback execution agency.

[0059] The data acquisition module is used to deploy a high-precision sensor network that integrates multiple types of sensors and communication units at key nodes in solid waste treatment to collect real-time data on solid waste types, quantities, equipment parameters and environmental data.

[0060] The digital twin engine is used to build a dynamic digital twin model based on LSTM and Simulink, and to predict its key parameters using the LSTM algorithm. Figure 3 The network structure of the digital twin engine submodule includes an input layer, a hidden layer, and an output layer; the input layer contains n features, namely S, Q, P, E The hidden layer is a 3-layer LSTM with 128, 64, and 32 neurons respectively; the output layer consists of 4 branches, corresponding to S, Q, P, E The predicted value of .

[0061] The optimization decision platform is used to deploy hybrid algorithm solvers and supports three-dimensional Pareto frontier visualization. By designing a multi-objective function of economy, environment and society, the Pareto optimal solution set is generated using a hybrid algorithm of NSGA-II and PSO. PSO is used to accelerate convergence through particle updates, while NSGA-II is used to ensure solution diversity through non-dominated sorting. The specific process of the hybrid optimization algorithm is as follows: Figure 4 As shown, the specific steps include: Initialize the population; The NSGA-II algorithm is used to generate the Pareto optimal solution set through non-dominated sorting; Calculate congestion; Determine whether the solution converges; if it converges, output the solution set; if it does not converge, execute the PSO algorithm based on the solid waste type, solid waste quantity, equipment operating parameters and environmental data predicted by the digital twin model, accelerate convergence through particle update collaboration, and jump to the step of generating the Pareto optimal solution set through non-dominated sorting using the NSGA-II algorithm.

[0062] The feedback actuator is used to dynamically adjust equipment parameters through the PLC controller. This part uses the feedback control module to adjust various operating parameters of the equipment in real time, ultimately achieving closed-loop management.

[0063] It should be pointed out that, according to the needs of implementation, the various steps / components described in this application can be split into more steps / components, or two or more steps / components or partial operations of steps / components can be combined into new steps / components to achieve the purpose of the present invention.

[0064] This embodiment also includes a processor, a communication interface, a memory and a communication bus; wherein the processor, the communication interface and the memory communicate with each other through the communication bus; the memory stores a computer program, and when the program is executed by the processor, the processor executes the steps of the multi-objective collaborative management method of industrial solid waste based on digital twins and hybrid optimization algorithms.

[0065] This embodiment also provides a computer-readable storage medium having executable instructions stored thereon. When the instructions are executed by a processor, the processor implements a multi-objective collaborative management method for industrial solid waste based on digital twins and hybrid optimization algorithms.

[0066] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware.

[0067] Moreover, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0068] This application is described with reference to the flowcharts of the method and computer program product according to Example 1 of the application and the block diagram of the device (system) according to Example 5. It should be understood that each process or block in the flowchart or block diagram, as well as combinations of processes or blocks in the flowchart or block diagram, can be implemented by computer program instructions.

[0069] These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce the instructions for implementing the process Figure 1 a process or multiple processes or boxes Figure 1 An industrial solid waste multi-objective collaborative management system based on digital twins and hybrid optimization algorithms with functions specified in a box or multiple boxes.

[0070] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes or boxes Figure 1 The function specified in one or more boxes.

[0071] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes or boxes Figure 1 The steps of the multi-objective collaborative management method of industrial solid waste based on digital twin and hybrid optimization algorithm are specified in one or more boxes.

[0072] The above embodiments are intended only to illustrate the design concepts and features of the present invention. Their purpose is to enable those skilled in the art to understand the contents of the present invention and implement them accordingly. The scope of protection of the present invention is not limited to the above embodiments. Therefore, any equivalent changes or modifications made based on the principles and design concepts disclosed in the present invention are within the scope of protection of the present invention.

Claims

1. A multi-objective collaborative management method for industrial solid waste based on digital twins and hybrid optimization algorithms, characterized by: The following steps are involved: S1: Collect current data including solid waste type, solid waste quantity, equipment operating parameters and environmental data through sensors, and obtain historical data through the database; S2: Build a digital twin model at the current moment based on current and historical data; use the LSTM algorithm to dynamically predict the type and quantity of solid waste, equipment operating parameters, and environmental data at the next moment and update the digital twin model; S3: Establish a multi-objective optimization function that includes economic, environmental, and social goals. Based on the next moment's solid waste type and quantity, equipment operating parameters, and environmental data output by the digital twin model, a hybrid algorithm of NSGA-II and PSO is used to solve the multi-objective optimization function and generate a Pareto optimal solution set. S4: Dynamically correct the digital twin model through the actual data collected by the sensor and the error function.

2. The multi-objective collaborative management method for industrial solid waste based on digital twin and hybrid optimization algorithm according to claim 1 is characterized by: In step S1, an Internet of Things sensor network is deployed in the industrial solid waste treatment process. The Internet of Things sensors include weight sensors, temperature and humidity sensors, GPS positioning modules and energy consumption monitoring units.

3. The multi-objective collaborative management method for industrial solid waste based on digital twin and hybrid optimization algorithm according to claim 1 is characterized by: In the step S3, The economic goal is to minimize the unit solid waste treatment cost and unit distance transportation cost while maximizing resource utilization benefits. Resource utilization benefits are the recycled resource benefits obtained per unit solid waste volume, which is equal to the product of the resource utilization benefit coefficient and the resource conversion efficiency. The unit solid waste treatment cost is calculated based on the predicted solid waste type and amount; the unit distance transportation cost is calculated based on the predicted solid waste amount; and the resource conversion efficiency is calculated based on the predicted solid waste type and equipment operating parameters. The environmental goal is to minimize carbon emissions and energy consumption. Carbon emissions and energy consumption are obtained by weighting the unit solid waste treatment cost and unit distance transportation cost using different environmental goal weight coefficients and carbon emission coefficients, respectively. The carbon emission coefficient is calculated based on the predicted equipment operating parameters and environmental data. The environmental goal weight coefficient corresponds to the carbon emission coefficient. The social goal is to maximize public satisfaction, which is equal to the weighted sum of resource conversion efficiency and carbon dioxide emission reduction according to the social goal weight coefficient; The amount of carbon dioxide emissions reduction is calculated in real time based on the predicted amount of solid waste and environmental data; the weight coefficient of social goals is dynamically adjusted through the entropy weight method.

4. The multi-objective collaborative management method for industrial solid waste based on digital twin and hybrid optimization algorithm according to claim 1 is characterized by: In step S3, the specific steps of using the hybrid algorithm of NSGA-II and PSO to solve the multi-objective optimization function are as follows: Initialize the population; The NSGA-II algorithm is used to generate the Pareto optimal solution set through non-dominated sorting; Calculate congestion; Determine whether the solution converges; if it converges, output the solution set; if it does not converge, execute the PSO algorithm based on the solid waste type, solid waste quantity, equipment operating parameters and environmental data predicted by the digital twin model, accelerate convergence through particle update collaboration, and jump to the step of generating the Pareto optimal solution set through non-dominated sorting using the NSGA-II algorithm.

5. The multi-objective collaborative management method for industrial solid waste based on digital twin and hybrid optimization algorithm according to claim 4 is characterized by: In step S3, the crossover probability and mutation probability of the NSGA-II algorithm in the hybrid algorithm are adaptively adjusted with the number of iterations.

6. The multi-objective collaborative management method for industrial solid waste based on digital twin and hybrid optimization algorithm according to claim 4 is characterized by: In step S3, the specific steps of using the PSO algorithm to accelerate convergence through particle update collaboration are as follows: Let the solution vector be the position of the particle at the current moment, and the velocity of the particle at the next iteration as the direction of change and step size of the solution; The particle's position at the current moment is subtracted from the individual cognition and social group positions respectively, and the two differences are weighted by the corresponding acceleration constant and random number respectively; then the sum is added with the particle's speed at the current moment weighted by the inertia weight to obtain the particle's speed at the next iteration.

7. The multi-objective collaborative management method for industrial solid waste based on digital twin and hybrid optimization algorithm according to claim 1 is characterized by: In step S4, the error function is the mean of the difference between the predicted value and the actual value output by the digital twin model in each round; when the error exceeds the threshold, the learning rate of the digital twin model is adjusted using the LSTM algorithm to update the learning rate to respond to emergencies.

8. A multi-objective collaborative management system for industrial solid waste based on digital twins and hybrid optimization algorithms, characterized by: The data acquisition submodule is used to collect current data including solid waste types, solid waste quantities, equipment operating parameters and environmental data through sensors, and obtain historical data through the database; The digital twin engine submodule is used to build a digital twin model at the current moment based on current and historical data. It uses the LSTM algorithm to dynamically predict the type and quantity of solid waste, equipment operating parameters, and environmental data at the next moment and update the digital twin model. Optimize the decision-making submodule and establish a multi-objective optimization function that includes economic goals, environmental goals, and social goals. Based on the solid waste type and quantity, equipment operating parameters, and environmental data at the next moment output by the digital twin model, a hybrid algorithm of NSGA-II and PSO is used to solve the multi-objective optimization function and generate a Pareto optimal solution set. The feedback correction submodule is used to dynamically correct the parameters of the digital twin model through the actual data collected by the sensor and the error function.

9. The multi-objective collaborative management system for industrial solid waste based on digital twin and hybrid optimization algorithm according to claim 8 is characterized by: The network structure of the digital twin engine submodule includes an input layer, a hidden layer, and an output layer; The input layer includes features such as the types of solid waste, the amount of solid waste, equipment operating parameters, and environmental data. The hidden layer is a multi-layer LSTM, which contains a doubling number of neurons; The output layer includes branches corresponding to the predicted values ​​of solid waste types, solid waste quantities, equipment operating parameters and environmental data.

10. A computer memory, characterized in that: It stores a computer program that can be executed by a computer processor, which executes the multi-objective collaborative management method of industrial solid waste based on digital twin and hybrid optimization algorithm as described in any one of claims 1 to 7.

Images