Distributed cold chain load and photovoltaic consumption multi-target cooperative regulation and control method and system

Abstract

The invention provides a distributed cold chain load and photovoltaic consumption multi-target cooperative regulation and control method and system, and the method comprises the steps: inputting the operation data of a cold chain load into a pre-constructed cold chain load time-space characteristic model, obtaining the fluctuation of the cold chain load, and carrying out the optimization with the minimization of the fluctuation of the cold chain load as a target, and obtaining a cold storage time window; the method comprises the following steps: inputting current operation data of cold chain-energy storage-photovoltaic into a pre-constructed cold chain-energy storage-photovoltaic multi-target aggregation regulation and control model, taking a cold storage time window as a constraint boundary, and adopting a third-generation non-dominated sorting genetic algorithm and a distributed alternating direction multiplier method for solving; a multi-target cooperative regulation and control result with the maximization of the photovoltaic consumption rate, the minimization of the operation cost and the optimization of the fresh-keeping quality of the agricultural products as targets is obtained; the economical efficiency, the reliability and the environmental protection property of the distributed energy system are remarkably improved, and an efficient, reliable and low-cost cooperative regulation and control scheme is provided for the cold chain energy system under the permeation of high-proportion renewable energy sources.

Description

Technical Field

[0001] The present invention relates to the field of photovoltaic absorption optimization operation and cold chain logistics energy optimization, and specifically to a distributed cold chain load and photovoltaic absorption multi-objective coordinated control method and system. Background Art

[0002] The coordinated optimization of distributed photovoltaic (PV) and cold chain loads has become a key direction for the low-carbon transformation of energy systems. However, distributed PV output is significantly affected by spatiotemporal factors such as sunlight intensity and cloud cover, while cold chain loads exhibit strong spatiotemporal heterogeneity. For example, the cooling demand of agricultural cold storage is related to the cargo turnover cycle, and cold chain loads in different geographical locations have different energy priorities. Existing centralized control methods rely on network-wide information exchange, making it difficult to dynamically respond to the local constraints of distributed resources, resulting in insufficient PV utilization and delayed load regulation.

[0003] Cold chain energy systems must simultaneously optimize photovoltaic utilization, operating costs, and agricultural product quality, but these objectives present nonlinear conflicts. For example, forcing energy storage charging to increase utilization could lead to overcharging risks, while over-reliance on cold chain load regulation could cause temperature fluctuations in storage, impacting agricultural product quality. Traditional multi-objective optimization algorithms, when dealing with high-dimensional constraints, face slow convergence and uneven distribution of Pareto solutions, making them difficult to meet the real-time requirements of hourly decision-making. Summary of the Invention

[0004] In order to solve the problems in the prior art where centralized control methods are difficult to dynamically respond to local constraints of distributed resources and where traditional multi-objective optimization algorithms face slow convergence and uneven distribution of solutions when dealing with multi-objective optimization problems, the present invention provides a multi-objective coordinated control method for distributed cold chain loads and photovoltaic consumption. The improvement lies in that the method comprises:

[0005] The cold chain load operation data is input into the pre-built cold chain load spatiotemporal characteristic model to obtain the cold chain load fluctuation. The cold storage time window is optimized with the goal of minimizing the cold chain load fluctuation.

[0006] The current operating data of the cold chain, energy storage, and photovoltaic system is input into a pre-built multi-objective aggregation control model of the cold chain, energy storage, and photovoltaic system. Using the cold storage time window as the constraint boundary, a third-generation non-dominated sorting genetic algorithm and a distributed alternating direction multiplier method are used to solve the problem. This results in a multi-objective coordinated control result with the goals of maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the quality of agricultural product preservation.

[0007] Among them, the spatiotemporal characteristic model of the cold chain load couples the time-space-cold storage characteristics, and the cold chain-energy storage-photovoltaic multi-objective aggregation control model is constructed with the goals of maximizing the photovoltaic absorption rate, minimizing operating costs and optimizing the preservation quality of agricultural products.

[0008] Preferably, inputting the cold chain load operation data into a pre-built cold chain load spatiotemporal characteristic model to obtain cold chain load fluctuations, and optimizing the cold storage time window with the goal of minimizing the cold chain load fluctuations includes:

[0009] The cold chain load operation data is input into the pre-built cold chain load spatiotemporal characteristic model. The multi-type resource constraints of the cold chain are used as constraints, and minimizing the cold chain load fluctuation is used as the optimization goal. The quadratic programming algorithm is used to solve the range of the cold storage time window.

[0010] Among them, the resource constraints of cold chain types include the cold chain load operation constraint matrix, the electric transport vehicle charging time and space constraint space, and the distributed photovoltaic output fluctuation constraint range.

[0011] Preferably, the expression of the cold chain load spatiotemporal characteristic model is:

[0012] L i (t) = α i (t)·β i ·[Q base,i +ΔQ i (t±Δt')]

[0013] Among them, L i (t) is the cold chain load value in period t, Q base,i is the basic cooling load of node i, ΔQ i (t±Δt') is the load time shift caused by cold storage / precooling operation, α i (t) is the time adjustment coefficient, β i is the spatial weight factor, t is the current period, and Δt' is the cooling storage time window;

[0014] Among them, the time adjustment coefficient is determined by temperature sensitivity and operation cycle, and the spatial weight factor is constructed through the hierarchical analysis method.

[0015] Preferably, the process of constructing the cold chain-energy storage-photovoltaic multi-objective aggregation control model includes:

[0016] With the goals of maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the quality of agricultural products, a photovoltaic absorption rate maximization function, an operating cost minimization function, and an agricultural product preservation quality optimization function were constructed. These photovoltaic absorption rate maximization function, operating cost minimization function, and agricultural product preservation quality optimization function were constrained by the allowable deviation range of cold chain storage temperature, the safe range of the energy storage system's state of charge, and the upper power load limit of the cold chain transportation path. This resulted in a multi-objective aggregation control model for cold chain, energy storage, and photovoltaics.

[0017] Among them, the photovoltaic absorption rate maximization function, the operating cost minimization function and the agricultural product preservation quality optimization function are each assigned different target weights.

[0018] Preferably, the expression of the photovoltaic absorption rate maximization function is:

[0019]

[0020] Among them, f1 is the photovoltaic absorption rate of the maximized period t, Ppv(t)=min(P PV (t),L cold (t)+P ess-ch (t)) is the photovoltaic power actually absorbed in period t, P PV (t) is the real-time photovoltaic output in period t, L cold (t) is the cold chain load demand in period t, P ess-ch (t) is the energy storage system charging power in time period t, and T represents the total duration of the evaluation period;

[0021] Preferably, the expression of the running cost minimization function is:

[0022]

[0023] Among them, f2 is the operating cost of the minimized period t, C ess is the energy storage charge and discharge cycle cost coefficient, P ess (t) is the energy storage charging and discharging power in period t, C grid (t) is the time-of-use electricity price in period t; P grid (t) is the power purchased by the grid during period t, C cold (t) is the operating cost of the cold chain equipment in period t, P cold (t) is the operating power of the cold chain equipment in period t, C OM is the annual maintenance cost of the equipment;

[0024] Preferably, the expression of the agricultural product preservation quality optimization function is:

[0025]

[0026] Among them, f3 is the optimal agricultural product preservation quality during period t, T act (τ) is the actual temperature of the cold chain storage at time period t, T set is the set temperature, Δt is the duration of the fresh-keeping cycle, j represents the jth fresh-keeping cycle, t j represents the starting time of the jth period, τ represents the integral variable of time period t, and J represents the total period.

[0027] Preferably, the current operating data of the cold chain-energy storage-photovoltaic system is input into a pre-built cold chain-energy storage-photovoltaic multi-objective aggregated control model, and the cold storage time window is used as the constraint boundary. The third-generation non-dominated sorting genetic algorithm and the distributed alternating direction multiplier method are used for solving the problem. The multi-objective coordinated control results with the goals of maximizing the photovoltaic absorption rate, minimizing the operating costs, and optimizing the freshness quality of agricultural products include:

[0028] The current operating data of the cold chain, energy storage, and photovoltaic system is collected, and the cold storage time window is used as the constraint boundary. Based on the current operating data, a multi-time scale rolling optimization is used on the first upper time scale using a third-generation non-dominated sorting genetic algorithm to solve the above multi-objective aggregation control model of the cold chain, energy storage, and photovoltaic system. The decision results are aimed at maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the freshness of agricultural products.

[0029] Based on the decision results, the distributed alternating direction multiplier method is used on the second time scale of the lower layer to minimize the total power deviation and the regulation cost optimization solution, and the final multi-objective coordinated control result is obtained;

[0030] Here, the first time scale is longer than the second time scale.

[0031] Preferably, the multi-time-scale rolling optimization based on the current operating data is used to solve the cold chain-energy storage-photovoltaic multi-objective aggregation control model using a third-generation non-dominated sorting genetic algorithm on the upper first time scale, and the decision results obtained with the goals of maximizing the photovoltaic absorption rate, minimizing the operating costs and optimizing the freshness quality of agricultural products include:

[0032] Based on current operating data, a multi-time-scale rolling optimization algorithm (3GNSGA) was used to solve the cold chain-energy storage-photovoltaic multi-objective aggregation control model on the upper first time scale. This algorithm obtained a set of non-inferior solutions, including photovoltaic absorption capacity, energy storage charging and discharging variables, and cold chain load, with the goals of maximizing photovoltaic absorption rate, minimizing operating costs, and optimizing agricultural product preservation quality, as well as target weights.

[0033] According to the changing trend of photovoltaic energy, fuzzy set theory is used to select the compromise solution as the decision result from the above non-inferior solution set through the normalized membership function.

[0034] Preferably, the distributed alternating direction multiplier method is used on the second time scale of the lower layer based on the decision result to minimize the total power deviation and the regulation cost optimization solution, and the final multi-objective coordinated control result includes:

[0035] Based on the decision results, the distributed alternating direction multiplier method is used on the second time scale of the lower layer to minimize the total power deviation and regulation cost as the optimization goal;

[0036] The optimization objective is decomposed into photovoltaic sub-problems, energy storage sub-problems, and cold chain sub-problems. By solving the photovoltaic sub-problems, energy storage sub-problems, and cold chain sub-problems, the photovoltaic output, energy storage charging and discharging power, and cold chain equipment operating power are obtained as local variables.

[0037] Summarize local variables, use the global power balance target as the global coordination variable, and use the target weight output by the upper layer as the initial multiplier;

[0038] Based on the global coordination variable and the multiplier, the local variables are independently updated, the global coordination variable is updated by summarizing the updated local variables, and the multiplier and penalty coefficient are updated according to the updated local variables and the global coordination variable;

[0039] The convergence of the algorithm is judged by the mutual residual and self-residual of the augmented Lagrangian. If the convergence condition is met, the iteration is stopped and the updated global coordination variable is obtained. Otherwise, the local variables and the global coordination variable are updated alternately, and then the multiplier is updated until the number of iterations reaches the maximum number of iterations and the loop is exited.

[0040] Among them, the photovoltaic sub-problem is to maximize the photovoltaic output without exceeding the maximum available power under the current lighting conditions; the energy storage sub-problem is to optimize the energy storage charging and discharging power while ensuring the safe operation of the energy storage system; the cold chain sub-problem is to adjust the cold chain load refrigeration power while ensuring that the warehouse temperature is within the allowable range.

[0041] Based on the same inventive concept, the present invention also provides a distributed cold chain load and photovoltaic consumption multi-objective coordinated control system, the system comprising:

[0042] The cold chain load spatiotemporal characteristic module is used to input the cold chain load operation data into the pre-built cold chain load spatiotemporal characteristic model to obtain the cold chain load fluctuation. The cold chain load fluctuation is minimized as the goal to optimize the cold storage time window.

[0043] The cold chain-energy storage-photovoltaic module is used to input the current operating data of the cold chain-energy storage-photovoltaic system into a pre-built cold chain-energy storage-photovoltaic multi-objective aggregation control model. Using the cold storage time window as the constraint boundary, it uses a third-generation non-dominated sorting genetic algorithm and a distributed alternating direction multiplier method to solve the problem. The model obtains a multi-objective coordinated control result with the goals of maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the freshness of agricultural products.

[0044] Among them, the spatiotemporal characteristic model of the cold chain load couples the time-space-cold storage characteristics, and the cold chain-energy storage-photovoltaic multi-objective aggregation control model is constructed with the goals of maximizing the photovoltaic absorption rate, minimizing operating costs and optimizing the preservation quality of agricultural products.

[0045] Preferably, the cold chain load spatiotemporal characteristic module inputs the cold chain load operation data into a pre-built cold chain load spatiotemporal characteristic model to obtain cold chain load fluctuations, and optimizes the cold storage time window with the goal of minimizing the cold chain load fluctuations, including:

[0046] The cold chain load operation data is input into the pre-built cold chain load spatiotemporal characteristic model. The multi-type resource constraints of the cold chain are used as constraints, and minimizing the cold chain load fluctuation is used as the optimization goal. The quadratic programming algorithm is used to solve the range of the cold storage time window.

[0047] Among them, the resource constraints of cold chain types include the cold chain load operation constraint matrix, the electric transport vehicle charging time and space constraint space, and the distributed photovoltaic output fluctuation constraint range.

[0048] Preferably, the expression of the cold chain load spatiotemporal characteristic model in the cold chain load spatiotemporal characteristic module is:

[0049] L i (t) = α i (t)·β i ·[Q base,i +ΔQ i (t±Δt')]

[0050] Among them, L i (t) is the cold chain load value in period t, Q base,i is the basic cooling load of node i, ΔQ i (t±Δt') is the load time shift caused by cold storage / precooling operation, α i (t) is the time adjustment coefficient, β i is the spatial weight factor, t is the current period, and Δt' is the cooling storage time window;

[0051] Among them, the time adjustment coefficient is determined by temperature sensitivity and operation cycle, and the spatial weight factor is constructed through the hierarchical analysis method.

[0052] Preferably, the process of constructing the cold chain-energy storage-photovoltaic multi-objective aggregation control model in the cold chain-energy storage-photovoltaic module includes:

[0053] With the goals of maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the quality of agricultural products, a photovoltaic absorption rate maximization function, an operating cost minimization function, and an agricultural product preservation quality optimization function were constructed. These photovoltaic absorption rate maximization function, operating cost minimization function, and agricultural product preservation quality optimization function were constrained by the allowable deviation range of cold chain storage temperature, the safe range of the energy storage system's state of charge, and the upper limit of the power carrying capacity of the cold chain transportation path, resulting in a cold chain-energy storage-photovoltaic multi-objective aggregation control model.

[0054] Among them, the photovoltaic absorption rate maximization function, the operating cost minimization function and the agricultural product preservation quality optimization function are each assigned different target weights.

[0055] Preferably, the expression of the photovoltaic absorption rate maximization function in the cold chain-energy storage-photovoltaic module is:

[0056]

[0057] Among them, f1 is the photovoltaic absorption rate of the maximized period t, Ppv(t)=min(P PV (t),L cold (t)+P ess-ch (t)) is the photovoltaic power actually absorbed in period t, P PV (t) is the real-time photovoltaic output in period t, L cold (t) is the cold chain load demand in period t, P ess-ch (t) is the energy storage system charging power in time period t, and T represents the total duration of the evaluation period;

[0058] The expression of the operating cost minimization function in the cold chain-energy storage-photovoltaic module is:

[0059]

[0060] Among them, f2 is the operating cost of the minimized period t, C ess is the energy storage charge and discharge cycle cost coefficient, P ess (t) is the energy storage charging and discharging power in period t, C grid (t) is the time-of-use electricity price in period t; P grid (t) is the power purchased by the grid during period t, C cold (t) is the operating cost of the cold chain equipment in period t, P cold (t) is the operating power of the cold chain equipment in period t, C OM is the annual maintenance cost of the equipment;

[0061] The expression of the optimization function for the quality of agricultural product preservation in the cold chain-energy storage-photovoltaic module is:

[0062]

[0063] Among them, f3 is the optimal agricultural product preservation quality during period t, T act (τ) is the actual temperature of the cold chain storage at time period t, T set is the set temperature, Δt is the duration of the fresh-keeping cycle, j represents the jth fresh-keeping cycle, t j represents the starting time of the jth period, τ represents the integral variable of time period t, and J represents the total period.

[0064] Preferably, the cold chain-energy storage-photovoltaic module inputs the current operating data of the cold chain-energy storage-photovoltaic system into a pre-built cold chain-energy storage-photovoltaic multi-objective aggregation control model, and uses the cold storage time window as the constraint boundary. The model is solved using the third-generation non-dominated sorting genetic algorithm and the distributed alternating direction multiplier method. The multi-objective coordinated control results obtained with the goals of maximizing the photovoltaic absorption rate, minimizing the operating costs, and optimizing the freshness quality of agricultural products include:

[0065] The current operating data of the cold chain, energy storage, and photovoltaic system is collected, and the cold storage time window is used as the constraint boundary. Based on the current operating data, a multi-time scale rolling optimization is used on the first upper time scale using a third-generation non-dominated sorting genetic algorithm to solve the above multi-objective aggregation control model of the cold chain, energy storage, and photovoltaic system. The decision results are aimed at maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the freshness of agricultural products.

[0066] Based on the decision results, the distributed alternating direction multiplier method is used on the second time scale of the lower layer to minimize the total power deviation and the regulation cost optimization solution, and the final multi-objective coordinated control result is obtained;

[0067] Here, the first time scale is longer than the second time scale.

[0068] Preferably, the cold chain-energy storage-photovoltaic module uses a multi-time scale rolling optimization based on current operating data to solve the cold chain-energy storage-photovoltaic multi-objective aggregation control model on the upper first time scale using a third-generation non-dominated sorting genetic algorithm, and obtains decision results with the goals of maximizing the photovoltaic absorption rate, minimizing the operating cost, and optimizing the freshness quality of agricultural products, including:

[0069] Based on current operating data, a multi-time-scale rolling optimization algorithm (3GNSGA) was used to solve the cold chain-energy storage-photovoltaic multi-objective aggregation control model on the upper first time scale. This algorithm obtained a set of non-inferior solutions, including photovoltaic absorption capacity, energy storage charging and discharging variables, and cold chain load, with the goals of maximizing photovoltaic absorption rate, minimizing operating costs, and optimizing agricultural product preservation quality, as well as target weights.

[0070] According to the changing trend of photovoltaic energy, fuzzy set theory is used to select the compromise solution as the decision result from the above non-inferior solution set through the normalized membership function.

[0071] Preferably, in the cold chain-energy storage-photovoltaic module, based on the decision result, a distributed alternating direction multiplier method is used on the lower second time scale to optimize and solve the minimization of total power deviation and regulation cost, and the final multi-objective coordinated control result includes:

[0072] Based on the decision results, the distributed alternating direction multiplier method is used on the second time scale of the lower layer to minimize the total power deviation and regulation cost as the optimization goal;

[0073] The optimization objective is decomposed into photovoltaic sub-problems, energy storage sub-problems, and cold chain sub-problems. By solving the photovoltaic sub-problems, energy storage sub-problems, and cold chain sub-problems, the photovoltaic output, energy storage charging and discharging power, and cold chain equipment operating power are obtained as local variables.

[0074] Summarize local variables, use the global power balance target as the global coordination variable, and use the target weight output by the upper layer as the initial multiplier;

[0075] Based on the global coordination variable and the multiplier, the local variables are independently updated, the global coordination variable is updated by summarizing the updated local variables, and the multiplier and penalty coefficient are updated according to the updated local variables and the global coordination variable;

[0076] The convergence of the algorithm is judged by the mutual residual and self-residual of the augmented Lagrangian. If the convergence condition is met, the iteration is stopped and the updated global coordination variable is obtained. Otherwise, the local variables and the global coordination variable are updated alternately, and then the multiplier is updated until the number of iterations reaches the maximum number of iterations and the loop is exited.

[0077] Among them, the photovoltaic sub-problem is to maximize the photovoltaic output without exceeding the maximum available power under the current lighting conditions; the energy storage sub-problem is to optimize the energy storage charging and discharging power while ensuring the safe operation of the energy storage system; the cold chain sub-problem is to adjust the cold chain load refrigeration power while ensuring that the warehouse temperature is within the allowable range.

[0078] In another aspect, the present application further provides a computing device comprising: at least one processor and a memory;

[0079] The memory is used to store one or more programs;

[0080] When the one or more programs are executed by the one or more processors, a multi-objective coordinated control method of distributed cold chain load and photovoltaic consumption as described above is implemented.

[0081] On the other hand, the present application also provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed, it implements a multi-objective coordinated control method of distributed cold chain load and photovoltaic absorption as described above.

[0082] Compared with the prior art, the present invention has the following beneficial effects:

[0083] The present invention provides a distributed cold chain load and photovoltaic absorption multi-objective coordinated control method and system, comprising: inputting the operating data of the cold chain load into a pre-constructed cold chain load spatiotemporal characteristic model to obtain the cold chain load fluctuation, and optimizing the cold chain load fluctuation as the goal to obtain the cold storage time window; inputting the current operating data of the cold chain-energy storage-photovoltaic into a pre-constructed cold chain-energy storage-photovoltaic multi-objective aggregation control model, and using the cold storage time window as the constraint boundary, adopting the third generation non-dominated sorting genetic algorithm and the distributed alternating direction multiplier method to solve, and obtaining the multi-objective coordinated control result with the goals of maximizing the photovoltaic absorption rate, minimizing the operating cost and optimizing the preservation quality of agricultural products; wherein, the cold chain load spatiotemporal characteristic model couples the time-space-cold storage characteristics, and the cold chain-energy storage-photovoltaic multi-objective aggregation The control model is constructed with the goals of maximizing the photovoltaic absorption rate, minimizing operating costs and optimizing the preservation quality of agricultural products; the cold chain load spatiotemporal characteristic model uses a weighted fusion method to quantify the elasticity of cold chain energy demand in different geographical locations, thereby improving the local photovoltaic absorption rate and reducing the impact of cold chain load regulation on the quality of agricultural products; the model optimized with the goals of maximizing the photovoltaic absorption rate, minimizing operating costs and optimizing the preservation quality of agricultural products adopts an improved third-generation non-dominated sorting genetic algorithm to solve the problem of uneven distribution of solution sets under high-dimensional constraints; through the core technological innovations of spatiotemporal matching evaluation, multi-objective aggregation optimization and hierarchical distributed collaboration, the economy, reliability and environmental protection of the distributed energy system have been significantly improved, providing an efficient, reliable and low-cost collaborative control solution for the cold chain energy system under a high proportion of renewable energy penetration. BRIEF DESCRIPTION OF THE DRAWINGS

[0084] Figure 1 This is a flow chart of a multi-objective coordinated control method for cold chain load and photovoltaic consumption provided by the present invention;

[0085] Figure 2 A step diagram of a multi-objective coordinated control method for cold chain load and photovoltaic consumption provided by the present invention;

[0086] Figure 3 This is a diagram of the architecture of a cold chain load spatiotemporal characteristic model provided by the present invention;

[0087] Figure 4 A schematic diagram of a time-scale rolling optimization principle of a layered architecture provided by the present invention;

[0088] Figure 5 A schematic diagram of a hierarchical optimization process based on an improved NSGA-III provided by the present invention;

[0089] Figure 6 A schematic diagram of reference points and reference lines provided by the present invention;

[0090] Figure 7This is a diagram of the multi-objective coordinated control system for cold chain load and photovoltaic consumption provided by the present invention;

[0091] Figure 8 The present invention provides an electronic device. DETAILED DESCRIPTION

[0092] This application aims to overcome the shortcomings of existing technologies and proposes a multi-objective coordinated control strategy for distributed cold chain load and photovoltaic consumption based on resource clustering. Its core objectives include:

[0093] 1) Build a distributed resource coordination mechanism with time and space matching

[0094] By extracting multiple resource constraints and employing a weighted fusion approach to quantify the elasticity of cold chain energy demand in different geographic locations, a spatiotemporal correlation between PV output and cold chain load is established. For example, cold chain warehouses near PV stations are assigned a higher spatial weight factor to prioritize the consumption of local surplus PV power. For cold storage facilities with longer cold storage cycles, their load adjustment window is extended to enable cross-time energy transfer. This mechanism can improve the local utilization rate of PV and reduce the impact of cold chain load regulation on agricultural product quality.

[0095] 2) Design a multi-objective aggregation optimization model and an efficient solution algorithm

[0096] A three-objective function is proposed: maximizing the photovoltaic integration rate, minimizing operating costs, and optimizing the quality of agricultural product preservation. Multi-level constraints are also introduced. For upper-level hourly decision-making, an improved NSGA-III algorithm is employed to address the uneven distribution of the Pareto front under high-dimensional constraints through adaptive reference point adjustment and hierarchical constraint dominance relationships. Furthermore, dynamic weighting rules are defined between these objectives, for example, prioritizing photovoltaic integration during periods of abundant sunlight while prioritizing cost optimization during peak load periods, enhancing the strategy's adaptability.

[0097] 3) Realize distributed collaborative control across time scales

[0098] For the five-minute corrections at the lower layer, a distributed power adjustment algorithm with spatiotemporal matching was designed based on the ADMM framework. Geographically proximal photovoltaic, energy storage, and cold chain units were divided into coordinated subgroups through resource clustering, with coordination variables preferentially exchanged within the subgroups. Furthermore, a hot start mechanism was introduced, using the multi-objective weights output by the upper layer NSGA-III as the initial multiplier for the lower layer ADMM. This ensured cross-layer policy consistency and avoided frequent power fluctuations.

[0099] This application significantly improves the economy, reliability and environmental protection of distributed energy systems through core technological innovations such as spatiotemporal matching evaluation, multi-objective aggregation optimization and hierarchical distributed collaboration, and provides key technical support for the high proportion of renewable energy consumption and the low-carbonization of the cold chain industry.

[0100] The specific embodiments of the present invention are further described in detail below with reference to the accompanying drawings.

[0101] Example 1

[0102] The present invention provides a multi-objective coordinated control method for cold chain load and photovoltaic consumption, such as Figure 1 ,include:

[0103] The cold chain load operation data is input into the pre-built cold chain load spatiotemporal characteristic model to obtain the cold chain load fluctuation. The cold storage time window is optimized with the goal of minimizing the cold chain load fluctuation.

[0104] The current operating data of the cold chain, energy storage, and photovoltaic system is input into a pre-built multi-objective aggregation control model of the cold chain, energy storage, and photovoltaic system. Using the cold storage time window as the constraint boundary, a third-generation non-dominated sorting genetic algorithm and a distributed alternating direction multiplier method are used to solve the problem. This results in a multi-objective coordinated control result with the goals of maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the quality of agricultural product preservation.

[0105] Among them, the spatiotemporal characteristic model of the cold chain load couples the time-space-cold storage characteristics, and the cold chain-energy storage-photovoltaic multi-objective aggregation control model is constructed with the goals of maximizing the photovoltaic absorption rate, minimizing operating costs and optimizing the preservation quality of agricultural products.

[0106] Preferably, the cold chain load operation data is input into a pre-built cold chain load spatiotemporal characteristic model to obtain the cold chain load fluctuation, and the cold storage time window is optimized with the goal of minimizing the cold chain load fluctuation, including:

[0107] The cold chain load operation data is input into the pre-built cold chain load spatiotemporal characteristic model. The multi-type resource constraints of the cold chain are used as constraints, and minimizing the cold chain load fluctuation is used as the optimization goal. The quadratic programming algorithm is used to solve the range of the cold storage time window.

[0108] Among them, the resource constraints of cold chain types include the cold chain load operation constraint matrix, the electric transport vehicle charging time and space constraint space, and the distributed photovoltaic output fluctuation constraint range.

[0109] Preferably, the expression of the cold chain load spatiotemporal characteristic model is:

[0110] L i (t) = α i (t)·β i ·[Q base,i +ΔQ i (t±Δt')]

[0111] Among them, L i(t) is the cold chain load value in period t, Q base,i is the basic cooling load of node i, ΔQ i (t±Δt') is the load time shift caused by cold storage / precooling operation, α i (t)∈[0,1] is the time adjustment coefficient, β i ∈R + is the spatial weight factor, t is the current period, Δt' is the cooling time window, R + is the range size coefficient of the space;

[0112] Among them, the time adjustment coefficient is determined by temperature sensitivity and operation cycle, and the spatial weight factor is constructed through the hierarchical analysis method.

[0113] Preferably, the construction process of the cold chain-energy storage-photovoltaic multi-objective aggregation control model includes:

[0114] With the goals of maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the quality of agricultural products, a photovoltaic absorption rate maximization function, an operating cost minimization function, and an agricultural product preservation quality optimization function were constructed. These photovoltaic absorption rate maximization function, operating cost minimization function, and agricultural product preservation quality optimization function were constrained by the allowable deviation range of cold chain storage temperature, the safe range of the energy storage system's state of charge, and the upper limit of the power carrying capacity of the cold chain transportation path, resulting in a cold chain-energy storage-photovoltaic multi-objective aggregation control model.

[0115] Among them, the photovoltaic absorption rate maximization function, the operating cost minimization function and the agricultural product preservation quality optimization function are each assigned different target weights.

[0116] Preferably, the expression of the photovoltaic absorption rate maximization function is:

[0117]

[0118] Among them, f1 is the photovoltaic absorption rate of the maximized period t, Ppv(t)=min(P PV (t),L cold (t)+P ess-ch (t)) is the photovoltaic power actually absorbed in period t, P PV (t) is the real-time photovoltaic output in period t, L cold (t) is the cold chain load demand in period t, P ess-ch (t) is the energy storage system charging power in time period t, and T represents the total duration of the evaluation period;

[0119] Preferably, the expression of the running cost minimization function is:

[0120]

[0121] Among them, f2 is the operating cost of the minimized period t, C ess is the energy storage charge and discharge cycle cost coefficient, P ess (t) is the energy storage charging and discharging power in period t, C grid (t) is the time-of-use electricity price in period t; P grid (t) is the power purchased by the grid during period t, C cold (t) is the operating cost of the cold chain equipment in period t, P cold (t) is the operating power of the cold chain equipment in period t, C OM is the annual maintenance cost of the equipment;

[0122] Preferably, the expression of the agricultural product preservation quality optimization function is:

[0123]

[0124] Among them, f3 is the optimal agricultural product preservation quality during period t, T act (τ) is the actual temperature of the cold chain storage at time period t, T set is the set temperature, Δt is the duration of the fresh-keeping cycle, j represents the jth fresh-keeping cycle, t j represents the starting time of the jth period, τ represents the integral variable of time period t, and J represents the total period.

[0125] Preferably, the current operating data of the cold chain-energy storage-photovoltaic system is input into a pre-built cold chain-energy storage-photovoltaic multi-objective aggregation control model, and the cold storage time window is used as the constraint boundary. The third-generation non-dominated sorting genetic algorithm and the distributed alternating direction multiplier method are used for solving the problem. The multi-objective coordinated control results with the goals of maximizing the photovoltaic absorption rate, minimizing the operating costs, and optimizing the freshness quality of agricultural products are obtained, including:

[0126] In order to achieve multi-time scale collaborative optimization and dynamic correction, a hierarchical optimization architecture is proposed, which includes an hourly multi-objective decision-making layer and a minute-level distributed correction layer. The system control accuracy and robustness are improved through spatiotemporal decoupling and rolling optimization.

[0127] The current operating data of the cold chain, energy storage, and photovoltaic system is collected, and the cold storage time window is used as the constraint boundary. Based on the current operating data, a multi-time scale rolling optimization is used on the first upper time scale using a third-generation non-dominated sorting genetic algorithm to solve the above multi-objective aggregation control model of the cold chain, energy storage, and photovoltaic system. The decision results are aimed at maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the freshness of agricultural products.

[0128] Based on the decision results, the distributed alternating direction multiplier method is used on the second time scale of the lower layer to minimize the total power deviation and the regulation cost optimization solution, and the final multi-objective coordinated control result is obtained;

[0129] The first time scale is longer than the second time scale, the first time scale is at the hour level, and the second time scale is at the minute level.

[0130] Preferably, based on the current operating data, a multi-time-scale rolling optimization is used to solve the cold chain-energy storage-photovoltaic multi-objective aggregation control model using a third-generation non-dominated sorting genetic algorithm on the upper first time scale, and the decision results obtained with the goals of maximizing the photovoltaic absorption rate, minimizing the operating costs, and optimizing the freshness quality of agricultural products include:

[0131] Based on current operating data, a multi-time-scale rolling optimization algorithm (3GNSGA) was used to solve the cold chain-energy storage-photovoltaic multi-objective aggregation control model on the upper first time scale. This algorithm obtained a set of non-inferior solutions, including photovoltaic absorption capacity, energy storage charging and discharging variables, and cold chain load, with the goals of maximizing photovoltaic absorption rate, minimizing operating costs, and optimizing agricultural product preservation quality, as well as target weights.

[0132] According to the changing trend of photovoltaic energy, fuzzy set theory is used to select the compromise solution as the decision result from the above non-inferior solution set through the normalized membership function.

[0133] Preferably, based on the decision result, the distributed alternating direction multiplier method is used on the second time scale of the lower layer to minimize the total power deviation and the regulation cost optimization solution, and the final multi-objective coordinated control result includes:

[0134] Based on the decision results, the distributed alternating direction multiplier method is used on the second time scale of the lower layer to minimize the total power deviation and regulation cost as the optimization goal;

[0135] The optimization objective is decomposed into photovoltaic sub-problems, energy storage sub-problems, and cold chain sub-problems. By solving the photovoltaic sub-problems, energy storage sub-problems, and cold chain sub-problems, the photovoltaic output, energy storage charging and discharging power, and cold chain equipment operating power are obtained as local variables.

[0136] Summarize local variables, use the global power balance target as the global coordination variable, and use the target weight output by the upper layer as the initial multiplier;

[0137] Based on the global coordination variable and the multiplier, the local variables are independently updated, the global coordination variable is updated by summarizing the updated local variables, and the multiplier and penalty coefficient are updated according to the updated local variables and the global coordination variable;

[0138] The convergence of the algorithm is judged by the mutual residual and self-residual of the augmented Lagrangian. If the convergence condition is met, the iteration is stopped and the updated global coordination variable is obtained. Otherwise, the local variables and the global coordination variable are updated alternately, and then the multiplier is updated until the number of iterations reaches the maximum number of iterations and the loop is exited.

[0139] Among them, the photovoltaic sub-problem is to maximize the photovoltaic output without exceeding the maximum available power under the current lighting conditions; the energy storage sub-problem is to optimize the energy storage charging and discharging power while ensuring the safe operation of the energy storage system; the cold chain sub-problem is to adjust the cold chain load refrigeration power while ensuring that the warehouse temperature is within the allowable range.

[0140] Compared with the prior art, the present invention has the following beneficial effects:

[0141] The distributed collaborative control strategy based on resource clustering proposed in this application has achieved remarkable results in photovoltaic consumption, economy and system stability through spatiotemporal matching optimization and hierarchical algorithm design: by constructing a spatiotemporal elasticity model of cold chain load and a resource clustering coordination mechanism, a precise match between photovoltaic output and cold chain demand is achieved. The improved NSGA-III algorithm, combined with dynamic weight rules, quickly generates high-precision Pareto solutions in hourly optimization, shortens the convergence time of the objective function, and reduces the operating cost compared to single-objective optimization. The lower-level correction algorithm based on ADMM compresses the convergence time of the 5-minute power deviation through resource clustering and hot start mechanism, reducing the adjustment cost loss rate.

[0142] This application provides an efficient, reliable, and low-cost coordinated control solution for cold chain energy systems under a high proportion of renewable energy penetration, which has significant economic and social benefits. It is suitable for multiple scenarios such as agricultural product cold chain and pharmaceutical warehousing, and has broad promotion value.

[0143] Example 2

[0144] Based on the same inventive concept, the present invention also provides an optimal embodiment of a multi-objective coordinated control method for distributed cold chain load and photovoltaic consumption, such as Figure 2 ,include:

[0145] S1: Extracting multi-type resource constraints, considering the time adjustment coefficient of cold chain load, spatial weight factor, and time shift window of cold storage and pre-cooling, characterizing cold chain energy demand in different geographical locations through weighted fusion method, and establishing a spatiotemporal matching evaluation system between distributed cold chain load and photovoltaic resources;

[0146] S2: Establish a three-objective optimization function that includes maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the quality of agricultural product preservation. Set multi-level constraints and build a cold chain-energy storage-photovoltaic multi-objective aggregation control model.

[0147] S3, designs a hierarchical optimization algorithm. The upper layer uses the NSGA-III algorithm to make hourly multi-objective optimization decisions. The lower layer uses ADMM distributed computing to achieve five-minute power correction and proposes a time-space matching local consumption coordinated control strategy.

[0148] Step S1 specifically includes:

[0149] S101: Extracting constraints for multiple types of cold chain resources

[0150] 1) Cold chain load operation constraint matrix

[0151] In response to the freshness requirements of different types of agricultural products, a temperature-time dual-dimensional constraint model is established. Temperature constraint: Set the temperature range allowed for each storage node [T min ,T max ], where T min is the minimum fresh-keeping temperature, T max The maximum energy consumption tolerance temperature. Time constraint: define the preservation time window [t start ,t end ], where t start is the starting point of the fresh-keeping time, t end The end point of the preservation time is characterized by the need for cold chain equipment to maintain temperature stability within a specified period of time.

[0152] Introducing the slack index

[0153]

[0154] Among them, T actual is the actual temperature, δ is the relaxation index, which measures the degree to which the actual temperature deviates from the theoretical lower limit. When δ>1, a temperature over-limit alarm is triggered, and when δ<0, it is judged as over-cooling.

[0155] Construct an N×M dimensional constraint matrix C cold , where: N is the number of cold storage nodes, M is the time granularity, and the matrix element c ij It represents the allowable temperature deviation threshold of the i-th node in the j-th period, and is combined with δ to generate the dynamic constraint boundary:

[0156]

[0157] 2) Temporal and spatial constraint model for charging electric transport vehicles

[0158] Optimize the charging time window of electric transport vehicles and define the charging time window coefficient:

[0159]

[0160] Among them, α is the charging time window coefficient, t act is the actual charging time, t opt During the peak period of photovoltaic output, charging behavior is prioritized to match it; Δt max is the maximum allowable charging time offset. When α<0.7, the charging period is determined to be infeasible.

[0161] Establish the charging power-time-location joint constraint space:

[0162] S ev ={P ev ,t,(x,y)} (4)

[0163] Among them, S ev represents the charging power-time-location joint constraint space, the charging power P ev Limited by the vehicle battery capacity and the rated power of the charging pile; the charging period t must meet the transportation task time window t task ±Δt buffer ;Δt buffer Indicates variable charging time, the charging location (x, y) must be located at the transport path node R path within the radiation range.

[0164] At the same time, the charging location constraints are updated according to real-time traffic information, and the Hungarian algorithm is used to solve the optimal matching between charging tasks and vehicle locations to ensure that charging behavior does not interrupt the transportation process.

[0165] 3) Extraction of distributed photovoltaic output fluctuation constraints

[0166] Calculate the PV power fluctuation rate on a 15-minute time scale:

[0167]

[0168] Where β is the photovoltaic power fluctuation rate, ΔP pv =P pv，t+1 -P pv，t is the power difference between adjacent time periods, P pv，t is the rated power of the photovoltaic power station in time period t, P pv，t+1 is the rated power of the photovoltaic power station in time period t+1, P PV is the rated power of the photovoltaic power station.

[0169] For the controllable range setting of distributed photovoltaics, when β≤30%, the fluctuation is considered to be within the controllable range and direct grid connection is allowed; when β>30%, the energy storage system or power reduction strategy needs to be activated.

[0170] S102: Aiming at the spatiotemporal heterogeneity of cold chain load, this paper proposes a three-dimensional modeling framework based on time adjustment coefficient, spatial weight factor and cold storage time window, such as Figure 3 The model uses cold storage nodes as basic units and achieves a refined representation of cold chain energy demand through dynamic coupling of time-space-cold storage characteristics.

[0171] Taking into account the time adjustment coefficient, spatial weight factor and time shift window of the cold chain load and the cold storage and pre-cooling link, a weighted fusion method is used to characterize the cold chain energy demand in different geographical locations and construct a spatiotemporal characteristic model of the cold chain load:

[0172] L i (t) = α i (t)·β i ·[Q base,i +ΔQ i (t±Δt')] (6)

[0173] Among them, L i (t) is the cold chain load value in period t, Q base,i is the basic cooling load of node i, ΔQ i (t±Δt') is the load time shift caused by cold storage / precooling operation, α i (t)∈[0,1] is the time adjustment coefficient, β i ∈R + is the spatial weight factor, t is the current period, Δt' is the cooling time window, R + is the range size coefficient of the space.

[0174] The time adjustment coefficient characterizes the dynamic characteristics of the cold chain load fluctuation over time, which is composed of the temperature sensitivity S T It is determined jointly by the operation cycle C(t).

[0175]

[0176] in, is the ambient temperature T out Load sensitivity due to changes, L i represents the cold storage load at the current moment, C(τ) represents the cold chain operation load at time τ; t0 represents the start time of the operation cycle; t represents the end time of the operation cycle; τ represents the integral variable, the time variable C(t) is the cold storage in and out operation cycle function, which is expanded using Fourier series, λ is the load inertia attenuation coefficient, a0 represents the basic cold chain operation volume, a n 、b n represents the Fourier coefficient, n represents the nth harmonic order, and k' represents the harmonic order.

[0177] The spatial weight factor is used to quantify the impact of geographical location on cold chain load, and an evaluation index system is constructed through the analytic hierarchy process (AHP).

[0178]

[0179] Among them, A i is the cold storage capacity of node i (unit: m 3 );D ijThe Euclidean distance between node i and adjacent node j (unit: km); N represents the total number of cold chain nodes in the region; A total Indicates the total cold storage capacity of the area (m 2 );R max Indicates the maximum regional circulation; R i It represents the circulation volume of regional agricultural products (unit: tons / day); w1 represents the weight coefficient of volume, w2 represents the weight coefficient of distance, and w3 represents the weight coefficient of circulation volume.

[0180] The cold storage time window Δt' is optimized. Considering the thermal inertia characteristics of the pre-cooling equipment, the cold storage time window must meet the following requirements:

[0181]

[0182] Among them, Q storage,i is the capacity of the cold storage device (kWh); η is the cold storage efficiency; is the cooling power, is the cooling power (kW).

[0183] With the goal of minimizing load fluctuation, the optimal Δt' is solved:

[0184]

[0185] Among them, L st,i is the load reference value, Δt' min is the minimum cold storage time window, Δt' max is the maximum cold storage time window, and the optimization model is solved using the quadratic programming algorithm.

[0186] S103: To achieve efficient coordination between photovoltaic power generation and cold chain loads, this paper proposes a two-dimensional evaluation system, encompassing both temporal matching and spatial coverage. This index system aims to quantitatively assess the coupling efficiency between photovoltaic output characteristics and the temporal and spatial distribution of cold chain loads, providing a theoretical basis for regional energy planning.

[0187] (1) Time dimension matching rate

[0188] The overlap between the photovoltaic output period and the cold chain load energy consumption period represents the temporal synchronization of energy supply and demand.

[0189]

[0190] Among them, η TMR is the overlap between the photovoltaic output period and the cold chain load energy consumption period, which represents the time synchronization of energy supply and demand. PV (t) is the photovoltaic output power in time period t, L c (t) is the cold chain load power in time period t, and T is the total duration of the evaluation period.

[0191] (2) Spatial dimension coverage

[0192] The proportion of cold chain load within the radiation radius of a photovoltaic power station reflects the spatial accessibility and economy of energy transportation.

[0193]

[0194] Among them, η SCD It represents the proportion of cold chain load within the radiation radius of the photovoltaic power station, reflecting the spatial accessibility and economy of energy transmission. c,i represents the power of the i-th cold chain load node; d i represents the Euclidean distance from load node i to the PV power station (km); R represents the economic radiation radius of the PV power station; N represents the number of cold chain nodes within the economic radiation radius of the PV power station; H(·) represents the Heaviside step function, H(x) = 1 (if x ≥ 0), otherwise 0; L ctotal is the total regional cold chain load.

[0195] By weighted fusion of temporal matching rate and spatial coverage, a comprehensive matching index is constructed:

[0196] η match =γ·η TMR +(1-γ)·η SCD

[0197] Among them, η match represents the comprehensive matching index, and γ represents the matching weight coefficient.

[0198] Step S2 specifically includes:

[0199] S201: For the cold chain-energy storage-photovoltaic system, a three-objective optimization function is established, including maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the quality of agricultural product preservation, to achieve coordinated optimization of energy utilization efficiency, economy, and agricultural product quality.

[0200] 1) Maximizing the photovoltaic absorption rate, the purpose of which is to maximize the direct utilization rate of photovoltaic power generation in the cold chain load and reduce the abandonment rate.

[0201]

[0202] Among them, f1 is the photovoltaic absorption rate of the maximized period t, Ppv(t)=min(P PV (t),L cold (t)+P ess-ch (t)) is the photovoltaic power actually absorbed in period t, P PV (t) is the real-time photovoltaic output in period t, L cold (t) is the cold chain load demand in period t, P ess-ch(t) is the charging power of the energy storage system in time period t, and T represents the total duration of the evaluation period.

[0203] The purpose is to maximize the direct utilization of photovoltaic power generation in cold chain loads and reduce the abandonment rate.

[0204] 2) Minimize operating costs, the purpose of which is to integrate energy storage loss costs, grid electricity purchase costs, cold chain operating costs and equipment operation and maintenance costs.

[0205]

[0206] Among them, f2 is the operating cost of the minimized period t, C ess is the energy storage charge and discharge cycle cost coefficient, P ess (t) is the energy storage charging and discharging power in period t, C grid (t) is the time-of-use electricity price in period t; P grid (t) is the power purchased by the grid during period t, C cold (t) is the operating cost of the cold chain equipment in period t, P cold (t) is the operating power of the cold chain equipment in period t, C OM is the annual maintenance cost of the equipment.

[0207] 3) Optimize the quality of agricultural product preservation. The purpose is to quantify the freshness through the cumulative temperature deviation index. The smaller the deviation, the higher the quality.

[0208]

[0209] Among them, f3 is the optimal agricultural product preservation quality during period t, T act (τ) is the actual temperature of the cold chain storage at time period t, T set is the set temperature, Δt is the duration of the fresh-keeping cycle, j represents the jth fresh-keeping cycle, t j represents the starting time of the jth period, τ represents the integral variable of time period t, and J represents the total period.

[0210] S202: For the cold chain-energy storage-PV system, set multi-level constraints, including the allowable deviation range of cold chain storage temperature, the safe range of the energy storage system's state of charge, and the upper limit of the cold chain transportation route's power capacity. Define the physical boundaries and control rules for the system's safe operation, including temperature stability, energy storage safety, and grid interaction restrictions.

[0211] 1) Allowable deviation of cold chain temperature

[0212] T set -△T≤T act (t)≤T set +△T (18)

[0213] Among them, Tset is the set temperature; ΔT is the temperature adjustment range.

[0214] 2) Safe operation of energy storage system

[0215]

[0216] Among them, P ESS,ch (t) represents the instantaneous value of charging power in time period t, P ESS,dis (t) represents the instantaneous value of discharge power in time period t, P ESS,out Indicates the maximum charge and discharge power threshold, B dis (t), B ch (t) is a mutually exclusive logical variable 0-1 variable, which must satisfy B dis (t)+B ch Exclusive constraint of (t)≤1.

[0217]

[0218] Among them, SOC min , SOC max The upper and lower thresholds of the allowed fluctuation range of the state of charge; SOC(t) is the real-time storage capacity of the energy storage in time period t; η ESS,ch ,η ESS,dis The energy conversion efficiency coefficients corresponding to the charging / discharging processes respectively.

[0219] 3) Power load limit of cold chain transport path

[0220] Ensure that the charging power of electric vehicles (EVs) does not exceed the combined limits of real-time available energy supply and equipment carrying capacity.

[0221] P ev (t)≤min(P pv-ev (t)+P ess-ev (t),P ev-max ) (twenty one)

[0222] Among them, P ev (t) represents the actual charging power of the electric vehicle at time t; P pv-ev (t) represents the power directly supplied to EV by photovoltaic power at time t; P ess-ev (t) represents the power that the energy storage system can provide at time t; P ev-max Indicates the maximum power carrying capacity of electric vehicles.

[0223] 4) Power balance

[0224] P pv (t)+P grid (t) = L cold (t)+P ess-ch (t)-Pess-dis (t) (22)

[0225] S203: Constructing a multi-objective aggregation control model for cold chain, energy storage and photovoltaics:

[0226] min[f1,f2,f3]^T

[0227] st-type 18-22 restraint (23)

[0228] Among them, the decision variable is the photovoltaic consumption capacity P pv (t), energy storage charge and discharge variable P ess (t) and cold chain load L cold (t).

[0229] Step S3 specifically includes:

[0230] S301: To achieve multi-timescale collaborative optimization and dynamic correction, a hierarchical optimization architecture is proposed, such as Figure 4 , including an hourly multi-objective decision-making layer and a minute-level distributed correction layer, which improves the system control accuracy and robustness through spatiotemporal decoupling and rolling optimization.

[0231] In the existing layered architecture, coordination between upper-level scheduling and lower-level corrections is insufficient. Upper-level hourly plans are typically based on static forecast data and fail to account for real-time fluctuations at the five-minute level. Meanwhile, lower-level distributed corrections lack dynamic feedback on the upper-level multi-objective weights, leading to increased regulation costs. Furthermore, traditional ADMM algorithms fail to fully utilize resource clustering in distributed iterations, resulting in high communication overhead and poor convergence stability.

[0232] This application solves the upper part of the optimization target based on the multi-objective evolutionary algorithm of the third generation non-dominated sorting genetic algorithm (NSGA-III), such as Figure 5 ,NSGA-III algorithm is the third generation non-dominated sorting genetic optimization algorithm for multi-objective optimization problems, which is formed by combining Pareto theory.

[0233] 1) Initialization. Read system information, including the current unit operating status, photovoltaic output information, etc. Use Latin Hypercube Sampling (LHS) to generate the initial population R o , ensuring that the decision variables are evenly distributed in the feasible region.

[0234] 2) Non-dominated solution screening and stratification. Multiple non-dominated layers are formed by fast non-dominated sorting, and the non-dominated layers F are placed into the new population S layer by layer. t , so that the population S t The number P reaches N, where N represents the number of reference points.

[0235] In order to ensure the diversity of solutions, NSGA-II uses the congestion ranking method to screen the last layer, but congestion ranking is quite difficult to deal with multi-objective problems. Therefore, the NSGA-III algorithm introduces widely distributed reference points, such as Figure 6 The essence is to divide the space into grids to judge the degree of crowding of each solution, and then select the solution with less crowding to ensure the diversity of the population.

[0236] 3) A multi-objective optimization screening strategy based on a preset reference point set is used to carry out screening at the last level by means of reference points and ensuring the wide distribution characteristics of the set reference points. First, a reference point set with uniform distribution characteristics is constructed in the target space, and then the objective function values ​​of the individuals in the population are standardized so that the values ​​of each dimension are mapped to the interval [0,1]. By establishing an association mechanism between candidate solutions and reference points, a reference vector (a directed line segment from a reference point to the origin of the coordinate system) is used to achieve a quantitative evaluation of the distribution density of the solution set. Specifically, the vertical distance between each candidate solution and its nearest reference vector in the standardized target space is calculated, and by counting the cardinality of the solution set associated with each reference point, individuals in sparse areas are preferentially retained to maintain population diversity. Figure 5 As shown in the figure, in the three-dimensional target space (F1, F2, F3), the spatial vector network composed of the reference point and the coordinate origin provides a quantitative basis for the evaluation of the solution set distribution, and each coordinate axis represents the dimensionless processing result of the independent optimization target.

[0237] Finally put S t Non-dominant layer F last Screening is performed and the selected individuals are placed in P t+1 In the t+1 The parent individuals of the generation form a new offspring population Q through crossover mutation operation t+1 , and terminate when the number of iterations Gen is equal to the maximum number of iterations.

[0238] P t+1 represents the parent population of the t+1th generation, F last represents the non-dominated frontier of the last layer, and K represents the objective function dimension (K=3 in this application).

[0239] 4) Compromise solution selection based on fuzzy logic. The best solution has the smallest distance to the optimal solution and the largest distance to the worst solution. This paper applies fuzzy set theory to select the compromise solution. The membership function is:

[0240]

[0241] Among them, u im is the membership function of the i-th solution and the m-th objective function; F i Indicates the value of the current target i-th solution; F i，max , Fi，min is the maximum and minimum value of the mth objective function. For minimizing the objective function, the minimum value is the optimal solution.

[0242] Then the normalized membership function of each solution is calculated as:

[0243]

[0244] Among them, u i for u im In the normalized membership, I is the set of non-dominated solutions and M is the set of objective functions.

[0245] The Pareto solution set of the three objective functions of the system is solved by the multi-objective evolutionary algorithm based on NSGA-III, and reasonable decisions are made in the Pareto solution set based on the changing trend of photovoltaic energy through fuzzy set theory.

[0246] S302: Based on the hourly decision-making results of the upper layer, distributed computing is used on a 5-minute timescale to correct power deviations and coordinate the real-time matching of photovoltaic, energy storage, and cold chain loads. An adaptive step-size ADMM optimization model is used to decompose the global optimization problem into three subproblems (photovoltaic, energy storage, and cold chain), alternating between updating local variables and global coordination variables, making it suitable for distributed collaborative optimization.

[0247] Minimize total power deviation and regulation cost:

[0248]

[0249] Among them, P plan is the global power balance target, ΔP cold It is the difference between the front and back that can be adjusted.

[0250] Total power deviation:

[0251] ΔP=P plan -(P pv +P ess +P cold ) (27)

[0252] Where ΔP is the total power deviation, P pv is the photovoltaic output, P ess is the energy storage charging and discharging power, P cold is the adjustment amount for cold chain load.

[0253] (1) Local optimization

[0254] Each unit is based on the current coordination variable ΔP (k) and multiplier λ (k) , solve the local subproblem:

[0255] 1) Photovoltaic sub-problems

[0256] Maximize output, constrained to the maximum available power under current lighting conditions:

[0257] max P pv

[0258] st0≤P pv ≤P pv，max (t) (28)

[0259] Among them, P pv，max (t) is the maximum rated power of the photovoltaic power station.

[0260] 2) Energy storage sub-problem

[0261] Optimize charging and discharging power to meet SOC safety range and charging and discharging rate:

[0262]

[0263] Among them, ρ represents the penalty coefficient, z (k) represents the global coordination variable, λ (k) Represents the multiplier, SOC represents the current state of charge of the energy storage system, SOC min Indicates the minimum state of charge of the energy storage system, SOC max Indicates the maximum state of charge of the energy storage system.

[0264] 3) Cold chain issues

[0265] Adjust the cooling power to ensure that the storage temperature is within the allowable range:

[0266]

[0267] Among them, T set Indicates setting temperature constraint, T min Indicates the minimum cold storage temperature, T max Indicates the maximum cold storage temperature.

[0268] (2) Coordinate variable updates

[0269] Summarize local results and update the global coordination variable z (k+1) :

[0270]

[0271] Among them, z (k+1) represents the updated global coordination variable, λ (k) represents the k-th generation multiplier, x i (k+1) represents the updated local solution of the subproblem, z represents the global coordination variable, and N represents the number of subproblems.

[0272] (3) Update of multipliers and penalty coefficients

[0273]

[0274] Among them, λ (k+1) represents the updated multiplier.

[0275]

[0276] Among them, ρ (k+1) represents the updated penalty coefficient, ρ (k) represents the penalty coefficient before update, ρ (k+1) represents the reference point vector, s (k+1) represents the actual point vector, and μ=10 is the empirical threshold.

[0277] Pass the latest coordination variables and updated multiplier z (k+1) ,λ (k+1) , calculate the mutual residuals and self-residuals of the augmented Lagrangian and judge the convergence of the algorithm.

[0278]

[0279] Among them, r es k+1 represents the original residual, s es k+1 represents the dual residual.

[0280] If the iteration termination condition (34) is met, the loop is exited; otherwise, the update is continued until the iteration termination condition is met or the maximum number of iterations is reached.

[0281]

[0282] Among them, ξ represents the threshold, k is the current iteration number, k max is the maximum number of iterations.

[0283] This application addresses the issues of photovoltaic absorption and cold chain load matching, and proposes a method based on the time adjustment coefficient and spatial weight factor of the cold chain load to characterize the geographical location and the distance to the photovoltaic station. Through the weighted fusion method, the spatiotemporal elasticity of resources is quantified, and the spatiotemporal association rules of distributed photovoltaic-energy storage-cold chain are established to realize the local resource priority coordination method.

[0284] This application constructs a three-objective optimization function, sets multi-level constraints, and establishes a multi-objective aggregate control model for cold chain, energy storage, and photovoltaics. A hierarchical optimization algorithm is proposed. The upper layer uses the NSGA-III algorithm for hourly multi-objective optimization decisions, while the lower layer uses ADMM distributed computing to achieve five-minute power corrections. A coordinated control strategy for local consumption with time-space matching is proposed.

[0285] Example 3

[0286] Based on the same inventive concept, the present invention also provides a distributed cold chain load and photovoltaic consumption multi-objective coordinated control system, such as Figure 7 , the system comprising:

[0287] The cold chain load spatiotemporal characteristic module is used to input the cold chain load operation data into the pre-built cold chain load spatiotemporal characteristic model to obtain the cold chain load fluctuation. The cold chain load fluctuation is minimized as the goal to optimize the cold storage time window.

[0288] The cold chain-energy storage-photovoltaic module is used to input the current operating data of the cold chain-energy storage-photovoltaic system into a pre-built cold chain-energy storage-photovoltaic multi-objective aggregation control model. Using the cold storage time window as the constraint boundary, it uses a third-generation non-dominated sorting genetic algorithm and a distributed alternating direction multiplier method to solve the problem. The model obtains a multi-objective coordinated control result with the goals of maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the freshness of agricultural products.

[0289] Among them, the spatiotemporal characteristic model of the cold chain load couples the time-space-cold storage characteristics, and the cold chain-energy storage-photovoltaic multi-objective aggregation control model is constructed with the goals of maximizing the photovoltaic absorption rate, minimizing operating costs and optimizing the preservation quality of agricultural products.

[0290] Preferably, the cold chain load spatiotemporal characteristic module inputs the cold chain load operation data into a pre-built cold chain load spatiotemporal characteristic model to obtain cold chain load fluctuations, and optimizes the cold storage time window with the goal of minimizing the cold chain load fluctuations, including:

[0291] The cold chain load operation data is input into the pre-built cold chain load spatiotemporal characteristic model. The multi-type resource constraints of the cold chain are used as constraints, and minimizing the cold chain load fluctuation is used as the optimization goal. The quadratic programming algorithm is used to solve the range of the cold storage time window.

[0292] Among them, the resource constraints of cold chain types include the cold chain load operation constraint matrix, the electric transport vehicle charging time and space constraint space, and the distributed photovoltaic output fluctuation constraint range.

[0293] Preferably, the expression of the cold chain load spatiotemporal characteristic model in the cold chain load spatiotemporal characteristic module is:

[0294] L i (t) = α i (t)·β i ·[Q base,i +ΔQ i (t±Δt')]

[0295] Among them, L i (t) is the cold chain load value in period t, Q base,i is the basic cooling load of node i, ΔQ i (t±Δt') is the load time shift caused by cold storage / precooling operation, α i (t) is the time adjustment coefficient, β i is the spatial weight factor, t is the current period, and Δt' is the cooling storage time window;

[0296] Among them, the time adjustment coefficient is determined by temperature sensitivity and operation cycle, and the spatial weight factor is constructed through the hierarchical analysis method.

[0297] Preferably, the process of constructing the cold chain-energy storage-photovoltaic multi-objective aggregation control model in the cold chain-energy storage-photovoltaic module includes:

[0298] With the goals of maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the quality of agricultural products, a photovoltaic absorption rate maximization function, an operating cost minimization function, and an agricultural product preservation quality optimization function were constructed. These photovoltaic absorption rate maximization function, operating cost minimization function, and agricultural product preservation quality optimization function were constrained by the allowable deviation range of cold chain storage temperature, the safe range of the energy storage system's state of charge, and the upper limit of the power carrying capacity of the cold chain transportation path, resulting in a cold chain-energy storage-photovoltaic multi-objective aggregation control model.

[0299] Among them, the photovoltaic absorption rate maximization function, the operating cost minimization function and the agricultural product preservation quality optimization function are each assigned different target weights.

[0300] Preferably, the expression of the photovoltaic absorption rate maximization function in the cold chain-energy storage-photovoltaic module is:

[0301]

[0302] Among them, f1 is the photovoltaic absorption rate of the maximized period t, Ppv(t)=min(P PV (t),L cold (t)+P ess-ch (t)) is the photovoltaic power actually absorbed in period t, P PV (t) is the real-time photovoltaic output in period t, L cold (t) is the cold chain load demand in period t, P ess-ch (t) is the energy storage system charging power in time period t, and T represents the total duration of the evaluation period;

[0303] The expression of the operating cost minimization function in the cold chain-energy storage-photovoltaic module is:

[0304]

[0305] Among them, f2 is the operating cost of the minimized period t, C ess is the energy storage charge and discharge cycle cost coefficient, P ess (t) is the energy storage charging and discharging power in period t, C grid (t) is the time-of-use electricity price in period t; P grid (t) is the power purchased by the grid during period t, C cold (t) is the operating cost of the cold chain equipment in period t, P cold (t) is the operating power of the cold chain equipment in period t, C OM is the annual maintenance cost of the equipment;

[0306] The expression of the optimization function for the quality of agricultural product preservation in the cold chain-energy storage-photovoltaic module is:

[0307]

[0308] Among them, f3 is the optimal agricultural product preservation quality during period t, T act (τ) is the actual temperature of the cold chain storage at time period t, T set is the set temperature, Δt is the duration of the fresh-keeping cycle, j represents the jth fresh-keeping cycle, t j represents the starting time of the jth period, τ represents the integral variable of time period t, and J represents the total period.

[0309] Preferably, the cold chain-energy storage-photovoltaic module inputs the current operating data of the cold chain-energy storage-photovoltaic system into a pre-built cold chain-energy storage-photovoltaic multi-objective aggregation control model, and uses the cold storage time window as the constraint boundary. The model is solved using the third-generation non-dominated sorting genetic algorithm and the distributed alternating direction multiplier method. The multi-objective coordinated control results obtained with the goals of maximizing the photovoltaic absorption rate, minimizing the operating costs, and optimizing the freshness quality of agricultural products include:

[0310] The current operating data of the cold chain, energy storage, and photovoltaic system is collected, and the cold storage time window is used as the constraint boundary. Based on the current operating data, a multi-time scale rolling optimization is used on the first upper time scale using a third-generation non-dominated sorting genetic algorithm to solve the above multi-objective aggregation control model of the cold chain, energy storage, and photovoltaic system. The decision results are aimed at maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the freshness of agricultural products.

[0311] Based on the decision results, the distributed alternating direction multiplier method is used on the second time scale of the lower layer to minimize the total power deviation and the regulation cost optimization solution, and the final multi-objective coordinated control result is obtained;

[0312] Here, the first time scale is longer than the second time scale.

[0313] Preferably, the cold chain-energy storage-photovoltaic module uses a multi-time scale rolling optimization based on current operating data to solve the cold chain-energy storage-photovoltaic multi-objective aggregation control model on the upper first time scale using a third-generation non-dominated sorting genetic algorithm, and obtains decision results with the goals of maximizing the photovoltaic absorption rate, minimizing the operating cost, and optimizing the freshness quality of agricultural products, including:

[0314] Based on current operating data, a multi-time-scale rolling optimization algorithm (3GNSGA) was used to solve the cold chain-energy storage-photovoltaic multi-objective aggregation control model on the upper first time scale. This algorithm obtained a set of non-inferior solutions, including photovoltaic absorption capacity, energy storage charging and discharging variables, and cold chain load, with the goals of maximizing photovoltaic absorption rate, minimizing operating costs, and optimizing agricultural product preservation quality, as well as target weights.

[0315] According to the changing trend of photovoltaic energy, fuzzy set theory is used to select the compromise solution as the decision result from the above non-inferior solution set through the normalized membership function.

[0316] Preferably, in the cold chain-energy storage-photovoltaic module, based on the decision result, a distributed alternating direction multiplier method is used on the lower second time scale to optimize and solve the minimization of total power deviation and regulation cost, and the final multi-objective coordinated control result includes:

[0317] Based on the decision results, the distributed alternating direction multiplier method is used on the second time scale of the lower layer to minimize the total power deviation and regulation cost as the optimization goal;

[0318] The optimization objective is decomposed into photovoltaic sub-problems, energy storage sub-problems, and cold chain sub-problems. By solving the photovoltaic sub-problems, energy storage sub-problems, and cold chain sub-problems, the photovoltaic output, energy storage charging and discharging power, and cold chain equipment operating power are obtained as local variables.

[0319] Summarize local variables, use the global power balance target as the global coordination variable, and use the target weight output by the upper layer as the initial multiplier;

[0320] Based on the global coordination variable and the multiplier, the local variables are independently updated, the global coordination variable is updated by summarizing the updated local variables, and the multiplier and penalty coefficient are updated according to the updated local variables and the global coordination variable;

[0321] The convergence of the algorithm is judged by the mutual residual and self-residual of the augmented Lagrangian. If the convergence condition is met, the iteration is stopped and the updated global coordination variable is obtained. Otherwise, the local variables and the global coordination variable are updated alternately, and then the multiplier is updated until the number of iterations reaches the maximum number of iterations and the loop is exited.

[0322] Among them, the photovoltaic sub-problem is to maximize the photovoltaic output without exceeding the maximum available power under the current lighting conditions; the energy storage sub-problem is to optimize the energy storage charging and discharging power while ensuring the safe operation of the energy storage system; the cold chain sub-problem is to adjust the cold chain load refrigeration power while ensuring that the warehouse temperature is within the allowable range.

[0323] Example 4

[0324] The present invention also provides an electronic device, such as Figure 8 As shown, the electronic device may be a computer, a single-chip microcomputer, a smart mobile device, etc. The electronic device in this embodiment may include a processor, a memory, a transceiver component, etc. The memory, processor, and transceiver component are connected via a bus; the memory may be used to store an execution program, which may include instructions; and the processor is used to execute the instructions stored in the memory. The memory may also be used to store data, which may be accessed and / or modified during the execution of the instructions.

[0325] The processor may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions in the storage medium to implement the corresponding method flow or corresponding function, so as to realize the steps of a distributed cold chain load and photovoltaic absorption multi-objective coordinated control method in the above embodiment.

[0326] Example 5

[0327] Based on the same inventive concept, the present invention also provides a readable storage medium, specifically an electronic device readable storage medium (Memory), which is a memory device in an electronic device for storing programs and data. It can be understood that the storage medium here can include both built-in storage media in the electronic device and, of course, extended storage media supported by the electronic device. The storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space. These instructions can be one or more execution programs (including program codes). It should be noted that the storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. The processor loads and executes one or more instructions stored in the storage medium, which can implement the steps of a multi-objective coordinated control method of distributed cold chain load and photovoltaic absorption in the above embodiment.

[0328] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0329] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. 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 generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0330] 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 and / or boxes Figure 1The function specified in one or more boxes.

[0331] 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 and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0332] The above are merely embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention are included in the scope of the claims of the present invention to be approved.

Claims

1. A multi-objective coordinated control method for distributed cold chain load and photovoltaic consumption, characterized in that: include: The cold chain load operation data is input into the pre-built cold chain load spatiotemporal characteristic model to obtain the cold chain load fluctuation. The cold storage time window is optimized with the goal of minimizing the cold chain load fluctuation. The current operating data of the cold chain, energy storage, and photovoltaic system is input into a pre-built multi-objective aggregation control model of the cold chain, energy storage, and photovoltaic system. Using the cold storage time window as the constraint boundary, a third-generation non-dominated sorting genetic algorithm and a distributed alternating direction multiplier method are used to solve the problem. This results in a multi-objective coordinated control result with the goals of maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the quality of agricultural product preservation. Among them, the spatiotemporal characteristic model of the cold chain load couples the time-space-cold storage characteristics, and the cold chain-energy storage-photovoltaic multi-objective aggregation control model is constructed with the goals of maximizing the photovoltaic absorption rate, minimizing operating costs and optimizing the preservation quality of agricultural products.

2. A multi-objective coordinated control method for distributed cold chain load and photovoltaic consumption according to claim 1, characterized in that: Inputting the cold chain load operation data into a pre-built cold chain load spatiotemporal characteristic model to obtain cold chain load fluctuations, and optimizing the cold storage time window with the goal of minimizing the cold chain load fluctuations includes: The cold chain load operation data is input into the pre-built cold chain load spatiotemporal characteristic model. The multi-type resource constraints of the cold chain are used as constraints, and minimizing the cold chain load fluctuation is used as the optimization goal. The quadratic programming algorithm is used to solve the range of the cold storage time window. Among them, the resource constraints of cold chain types include the cold chain load operation constraint matrix, the electric transport vehicle charging time and space constraint space, and the distributed photovoltaic output fluctuation constraint range.

3. A distributed cold chain load and photovoltaic consumption multi-objective coordinated control method according to claim 2, characterized in that: The expression of the spatiotemporal characteristic model of the cold chain load is: L i (t)=α i (t)·β i ·[Q base,i +ΔQ i (t±Δt')] Among them, L i (t) is the cold chain load value in period t, Q base,i is the basic cooling load of node i, ΔQ i (t±Δt') is the load time shift caused by cold storage / precooling operation, α i (t) is the time adjustment coefficient, β i is the spatial weight factor, t is the current period, and Δt' is the cooling storage time window; Among them, the time adjustment coefficient is determined by temperature sensitivity and operation cycle, and the spatial weight factor is constructed through the hierarchical analysis method.

4. A multi-objective coordinated control method for distributed cold chain load and photovoltaic consumption according to claim 1, characterized in that: The construction process of the cold chain-energy storage-photovoltaic multi-objective aggregation control model includes: With the goals of maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the quality of agricultural products, a photovoltaic absorption rate maximization function, an operating cost minimization function, and an agricultural product preservation quality optimization function were constructed. These photovoltaic absorption rate maximization function, operating cost minimization function, and agricultural product preservation quality optimization function were constrained by the allowable deviation range of cold chain storage temperature, the safe range of the energy storage system's state of charge, and the upper power load limit of the cold chain transportation path. This resulted in a multi-objective aggregation control model for cold chain, energy storage, and photovoltaics. Among them, the photovoltaic absorption rate maximization function, the operating cost minimization function and the agricultural product preservation quality optimization function are each assigned different target weights.

5. A multi-objective coordinated control method for distributed cold chain load and photovoltaic consumption according to claim 4, characterized in that: The expression of the photovoltaic absorption rate maximization function is: Among them, f1 is the photovoltaic absorption rate of the maximized period t, Ppv(t)=min(P PV (t),L cold (t)+P ess-ch (t)) is the photovoltaic power actually absorbed in period t, P PV (t) is the real-time photovoltaic output in period t, L cold (t) is the cold chain load demand in period t, P ess-ch (t) is the energy storage system charging power in time period t, and T represents the total duration of the evaluation period; The expression of the running cost minimization function is: Among them, f2 is the operating cost of the minimized period t, C ess is the energy storage charge and discharge cycle cost coefficient, P ess (t) is the energy storage charging and discharging power in period t, C grid (t) is the time-of-use electricity price in period t; P grid (t) is the power purchased by the grid during period t, C cold (t) is the operating cost of the cold chain equipment in period t, P cold (t) is the operating power of the cold chain equipment in period t, C OM is the annual maintenance cost of the equipment; The expression of the agricultural product preservation quality optimization function is: Among them, f3 is the optimal agricultural product preservation quality during period t, T act (τ) is the actual temperature of the cold chain storage at time period t, T set is the set temperature, Δt is the duration of the fresh-keeping cycle, j represents the jth fresh-keeping cycle, t j represents the starting time of the jth period, τ represents the integral variable of time period t, and J represents the total period.

6. A multi-objective coordinated control method for distributed cold chain load and photovoltaic consumption according to claim 5, characterized in that: The current operating data of the cold chain, energy storage, and photovoltaics are input into a pre-built cold chain, energy storage, and photovoltaic multi-objective aggregation control model. The cold storage time window is used as the constraint boundary, and the third-generation non-dominated sorting genetic algorithm and the distributed alternating direction multiplier method are used to solve the problem. The multi-objective coordinated control results with the goals of maximizing the photovoltaic absorption rate, minimizing the operating costs, and optimizing the freshness of agricultural products are obtained, including: The current operating data of the cold chain, energy storage, and photovoltaic system is collected, and the cold storage time window is used as the constraint boundary. Based on the current operating data, a multi-time scale rolling optimization is used on the first upper time scale using a third-generation non-dominated sorting genetic algorithm to solve the above multi-objective aggregation control model of the cold chain, energy storage, and photovoltaic system. The decision results are aimed at maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the freshness of agricultural products. Based on the decision results, the distributed alternating direction multiplier method is used on the second time scale of the lower layer to minimize the total power deviation and the regulation cost optimization solution, and the final multi-objective coordinated control result is obtained; Here, the first time scale is longer than the second time scale.

7. A multi-objective coordinated control method for distributed cold chain load and photovoltaic consumption according to claim 6, characterized in that: The above-mentioned cold chain-energy storage-photovoltaic multi-objective aggregation control model is solved by using a multi-time-scale rolling optimization method based on the current operating data at the upper first time scale using a third-generation non-dominated sorting genetic algorithm. The decision results obtained with the goals of maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the quality of agricultural product preservation include: Based on current operating data, a multi-time-scale rolling optimization algorithm (3GNSGA) was used to solve the cold chain-energy storage-photovoltaic multi-objective aggregation control model on the upper first time scale. This algorithm obtained a set of non-inferior solutions, including photovoltaic absorption capacity, energy storage charging and discharging variables, and cold chain load, with the goals of maximizing photovoltaic absorption rate, minimizing operating costs, and optimizing agricultural product preservation quality, as well as target weights. According to the changing trend of photovoltaic energy, fuzzy set theory is used to select the compromise solution as the decision result from the above non-inferior solution set through the normalized membership function.

8. A multi-objective coordinated control method for distributed cold chain load and photovoltaic consumption according to claim 7, characterized in that: Based on the decision result, the distributed alternating direction multiplier method is used on the second time scale of the lower layer to minimize the total power deviation and the regulation cost optimization solution, and the final multi-objective coordinated control result includes: Based on the decision results, the distributed alternating direction multiplier method is used on the second time scale of the lower layer to minimize the total power deviation and regulation cost as the optimization goal; The optimization objective is decomposed into photovoltaic sub-problems, energy storage sub-problems, and cold chain sub-problems. By solving the photovoltaic sub-problems, energy storage sub-problems, and cold chain sub-problems, the photovoltaic output, energy storage charging and discharging power, and cold chain equipment operating power are obtained as local variables. Summarize local variables, use the global power balance target as the global coordination variable, and use the target weight output by the upper layer as the initial multiplier; Based on the global coordination variable and the multiplier, the local variables are independently updated, the global coordination variable is updated by summarizing the updated local variables, and the multiplier and penalty coefficient are updated according to the updated local variables and the global coordination variable; The convergence of the algorithm is judged by the mutual residual and self-residual of the augmented Lagrangian. If the convergence condition is met, the iteration is stopped and the updated global coordination variable is obtained. Otherwise, the local variables and the global coordination variable are updated alternately, and then the multiplier is updated until the number of iterations reaches the maximum number of iterations and the loop is exited. Among them, the photovoltaic sub-problem is to maximize the photovoltaic output without exceeding the maximum available power under the current lighting conditions; the energy storage sub-problem is to optimize the energy storage charging and discharging power while ensuring the safe operation of the energy storage system; the cold chain sub-problem is to adjust the cold chain load refrigeration power while ensuring that the warehouse temperature is within the allowable range.

9. A distributed cold chain load and photovoltaic consumption multi-objective coordinated control system, characterized in that: include: The cold chain load spatiotemporal characteristic module is used to input the cold chain load operation data into the pre-built cold chain load spatiotemporal characteristic model to obtain the cold chain load fluctuation. The cold chain load fluctuation is minimized as the goal to optimize the cold storage time window. The cold chain-energy storage-photovoltaic module is used to input the current operating data of the cold chain-energy storage-photovoltaic system into a pre-built cold chain-energy storage-photovoltaic multi-objective aggregation control model. Using the cold storage time window as the constraint boundary, it uses a third-generation non-dominated sorting genetic algorithm and a distributed alternating direction multiplier method to solve the problem. The model obtains a multi-objective coordinated control result with the goals of maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the freshness of agricultural products. Among them, the spatiotemporal characteristic model of the cold chain load couples the time-space-cold storage characteristics, and the cold chain-energy storage-photovoltaic multi-objective aggregation control model is constructed with the goals of maximizing the photovoltaic absorption rate, minimizing operating costs and optimizing the preservation quality of agricultural products.

10. A distributed cold chain load and photovoltaic consumption multi-objective coordinated control system according to claim 9, characterized in that: The cold chain load spatiotemporal characteristic module inputs the cold chain load operation data into a pre-built cold chain load spatiotemporal characteristic model to obtain the cold chain load fluctuation. The cold storage time window is optimized with the goal of minimizing the cold chain load fluctuation. The cold chain load fluctuation is as follows: The cold chain load operation data is input into the pre-built cold chain load spatiotemporal characteristic model. The multi-type resource constraints of the cold chain are used as constraints, and minimizing the cold chain load fluctuation is used as the optimization goal. The quadratic programming algorithm is used to solve the range of the cold storage time window. Among them, the resource constraints of cold chain types include the cold chain load operation constraint matrix, the electric transport vehicle charging time and space constraint space, and the distributed photovoltaic output fluctuation constraint range.

11. A distributed cold chain load and photovoltaic consumption multi-objective coordinated control system according to claim 10, characterized in that: The expression of the cold chain load spatiotemporal characteristic model in the cold chain load spatiotemporal characteristic module is: L i (t)=α i (t)·β i ·[Q base,i +ΔQ i (t±Δt')] Among them, L i (t) is the cold chain load value in period t, Q base,i is the basic cooling load of node i, ΔQ i (t±Δt') is the load time shift caused by cold storage / precooling operation, α i (t) is the time adjustment coefficient, β i is the spatial weight factor, t is the current period, and Δt' is the cooling storage time window; Among them, the time adjustment coefficient is determined by temperature sensitivity and operation cycle, and the spatial weight factor is constructed through the hierarchical analysis method.

12. A distributed cold chain load and photovoltaic consumption multi-objective coordinated control system according to claim 9, characterized in that: The process of constructing the cold chain-energy storage-photovoltaic multi-objective aggregation control model in the cold chain-energy storage-photovoltaic module includes: With the goals of maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the quality of agricultural products, a photovoltaic absorption rate maximization function, an operating cost minimization function, and an agricultural product preservation quality optimization function were constructed. These photovoltaic absorption rate maximization function, operating cost minimization function, and agricultural product preservation quality optimization function were constrained by the allowable deviation range of cold chain storage temperature, the safe range of the energy storage system's state of charge, and the upper limit of the power carrying capacity of the cold chain transportation path, resulting in a cold chain-energy storage-photovoltaic multi-objective aggregation control model. Among them, the photovoltaic absorption rate maximization function, the operating cost minimization function and the agricultural product preservation quality optimization function are each assigned different target weights.

13. A distributed cold chain load and photovoltaic consumption multi-objective coordinated control system according to claim 12, characterized in that: The expression of the photovoltaic absorption rate maximization function in the cold chain-energy storage-photovoltaic module is: Among them, f1 is the photovoltaic absorption rate of the maximized period t, Ppv(t)=min(P PV (t),L cold (t)+P ess-ch (t)) is the photovoltaic power actually absorbed in period t, P PV (t) is the real-time photovoltaic output in period t, L cold (t) is the cold chain load demand in period t, P ess-ch (t) is the energy storage system charging power in time period t, and T represents the total duration of the evaluation period; The expression of the operating cost minimization function in the cold chain-energy storage-photovoltaic module is: Among them, f2 is the operating cost of the minimized period t, C ess is the energy storage charge and discharge cycle cost coefficient, P ess (t) is the energy storage charging and discharging power in period t, C grid (t) is the time-of-use electricity price in period t; P grid (t) is the power purchased by the grid during period t, C cold (t) is the operating cost of the cold chain equipment in period t, P cold (t) is the operating power of the cold chain equipment in period t, C OM is the annual maintenance cost of the equipment; The expression of the optimization function for the quality of agricultural product preservation in the cold chain-energy storage-photovoltaic module is: Among them, f3 is the optimal agricultural product preservation quality during period t, T act (τ) is the actual temperature of the cold chain storage at time period t, T set is the set temperature, Δt is the duration of the fresh-keeping cycle, j represents the jth fresh-keeping cycle, t j represents the starting time of the jth period, τ represents the integral variable of time period t, and J represents the total period.

14. A distributed cold chain load and photovoltaic consumption multi-objective coordinated control system according to claim 13, characterized in that: The cold chain-energy storage-photovoltaic module inputs the current operating data of the cold chain-energy storage-photovoltaic into a pre-built cold chain-energy storage-photovoltaic multi-objective aggregation control model. Using the cold storage time window as the constraint boundary, the model is solved using a third-generation non-dominated sorting genetic algorithm and a distributed alternating direction multiplier method. The multi-objective collaborative control results obtained, which aim to maximize the photovoltaic absorption rate, minimize operating costs, and optimize the quality of agricultural product preservation, include: The current operating data of the cold chain, energy storage, and photovoltaic system is collected, and the cold storage time window is used as the constraint boundary. Based on the current operating data, a multi-time scale rolling optimization is used on the first upper time scale using a third-generation non-dominated sorting genetic algorithm to solve the above multi-objective aggregation control model of the cold chain, energy storage, and photovoltaic system. The decision results are aimed at maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the freshness of agricultural products. Based on the decision results, the distributed alternating direction multiplier method is used on the second time scale of the lower layer to minimize the total power deviation and the regulation cost optimization solution, and the final multi-objective coordinated control result is obtained; Here, the first time scale is longer than the second time scale.

15. A distributed cold chain load and photovoltaic consumption multi-objective coordinated control system according to claim 14, characterized in that: In the cold chain-energy storage-photovoltaic module, based on current operating data, a multi-time scale rolling optimization is used to solve the cold chain-energy storage-photovoltaic multi-objective aggregation control model using a third-generation non-dominated sorting genetic algorithm on the first upper time scale. The decision results obtained with the goals of maximizing the photovoltaic absorption rate, minimizing operating costs, and optimizing the freshness of agricultural products include: Based on current operating data, a multi-time-scale rolling optimization algorithm (3GNSGA) was used to solve the cold chain-energy storage-photovoltaic multi-objective aggregation control model on the upper first time scale. This algorithm obtained a set of non-inferior solutions, including photovoltaic absorption capacity, energy storage charging and discharging variables, and cold chain load, with the goals of maximizing photovoltaic absorption rate, minimizing operating costs, and optimizing agricultural product preservation quality, as well as target weights. According to the changing trend of photovoltaic energy, fuzzy set theory is used to select the compromise solution as the decision result from the above non-inferior solution set through the normalized membership function.

16. A distributed cold chain load and photovoltaic consumption multi-objective coordinated control system according to claim 15, characterized in that: In the cold chain-energy storage-photovoltaic module, based on the decision results, a distributed alternating direction multiplier method is used on the lower second time scale to optimize and solve the minimization of total power deviation and regulation cost. The final multi-objective coordinated control results include: Based on the decision results, the distributed alternating direction multiplier method is used on the second time scale of the lower layer to minimize the total power deviation and regulation cost as the optimization goal; The optimization objective is decomposed into photovoltaic sub-problems, energy storage sub-problems, and cold chain sub-problems. By solving the photovoltaic sub-problems, energy storage sub-problems, and cold chain sub-problems, the photovoltaic output, energy storage charging and discharging power, and cold chain equipment operating power are obtained as local variables. Summarize local variables, use the global power balance target as the global coordination variable, and use the target weight output by the upper layer as the initial multiplier; Based on the global coordination variable and the multiplier, the local variables are independently updated, the global coordination variable is updated by summarizing the updated local variables, and the multiplier and penalty coefficient are updated according to the updated local variables and the global coordination variable; The convergence of the algorithm is judged by the mutual residual and self-residual of the augmented Lagrangian. If the convergence condition is met, the iteration is stopped and the updated global coordination variable is obtained. Otherwise, the local variables and the global coordination variable are updated alternately, and then the multiplier is updated until the number of iterations reaches the maximum number of iterations and the loop is exited. Among them, the photovoltaic sub-problem is to maximize the photovoltaic output without exceeding the maximum available power under the current lighting conditions; the energy storage sub-problem is to optimize the energy storage charging and discharging power while ensuring the safe operation of the energy storage system; the cold chain sub-problem is to adjust the cold chain load refrigeration power while ensuring that the warehouse temperature is within the allowable range.

17. A computer device, characterized in that: include: at least one processor and memory; The memory and the processor are connected via a bus; The memory is used to store one or more programs; When the one or more programs are executed by the at least one processor, a multi-objective coordinated control method of distributed cold chain load and photovoltaic consumption as described in any one of claims 1 to 8 is implemented.

18. A computer-readable storage medium, characterized in that An execution program is stored thereon, and when the execution program is executed, a multi-objective coordinated control method of distributed cold chain load and photovoltaic consumption as described in any one of claims 1 to 8 is implemented.

Images