Energy management and recovery method and system for cold chain supply chain

Abstract

The invention relates to the technical field of cold-chain supply chains, and discloses an energy management and recovery method and system for a cold-chain supply chain. The method comprises the following steps: acquiring and preprocessing operating parameters and environmental parameters of the cold chain system to obtain multi-dimensional feature data; performing feature extraction and local density distribution calculation on the multi-dimensional feature data to obtain energy consumption mode clustering data; constructing an equipment operation knowledge graph according to the energy consumption mode clustering data; performing abnormal feature analysis on the equipment operation knowledge graph and the energy consumption mode clustering data through a hybrid neural network model to obtain equipment energy consumption abnormal feature data; generating an initial energy recovery scheme according to the equipment energy consumption abnormal feature data; the initial energy recovery scheme is applied to the cold chain system, the operation efficiency data and the load matching data are collected, online iterative optimization is carried out, the target energy management scheme is obtained, dynamic adjustment and continuous optimization of the energy recovery scheme are achieved, and the energy utilization efficiency of the cold chain system is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of cold chain supply chain, and in particular to an energy management recovery method and system for a cold chain supply chain. Background Art

[0002] With the rapid development of the cold chain logistics industry, the scale of cold chain facilities continues to expand, energy consumption continues to increase, and the problem of high energy consumption has become increasingly prominent. The energy management method of the traditional cold chain system mainly relies on manual experience for adjustment, lacks systematic energy recovery and optimization strategies, resulting in a large amount of waste heat energy being directly discharged into the environment, causing energy waste.

[0003] At present, the energy management of cold chain systems generally has problems such as difficulty in accurately controlling equipment operating parameters, low energy recovery efficiency, and poor system load matching. Since the cold chain system involves multiple core components such as compressor units, condensers, and evaporators, there are complex energy transfer relationships and thermal coupling characteristics between the components. The traditional single control strategy is difficult to achieve system-level energy optimization management, which restricts the improvement of the overall energy efficiency of the cold chain system. In addition, the multi-source heterogeneous data generated during the operation of the cold chain system has not been fully utilized, and there is a lack of intelligent data analysis and decision support methods. Problems such as untimely monitoring of system operating status, delayed identification of abnormal operating conditions, and insufficient optimization of energy recovery solutions have led to unsatisfactory energy management results, making it difficult to adapt to the urgent needs of modern cold chain logistics for energy conservation and consumption reduction. Summary of the invention

[0004] The present invention provides an energy management recovery method and system for a cold chain supply chain, which are used to achieve dynamic adjustment and continuous optimization of energy recovery schemes and improve the energy utilization efficiency of the cold chain system.

[0005] In a first aspect, the present invention provides an energy management and recovery method for a cold chain supply chain, the energy management and recovery method for a cold chain supply chain comprising: Collect and preprocess the operating parameters and environmental parameters of the cold chain system to obtain multi-dimensional feature data; Performing feature extraction and local density distribution calculation on the multi-dimensional feature data to obtain energy consumption pattern clustering data; Constructing a device operation knowledge graph based on the energy consumption pattern clustering data; Performing abnormal feature analysis on the equipment operation knowledge graph and the energy consumption pattern clustering data through a hybrid neural network model to obtain abnormal feature data of equipment energy consumption; Generate an initial energy recovery plan based on the abnormal energy consumption characteristic data of the equipment; The initial energy recovery scheme is applied to the cold chain system, operation efficiency data and load matching data are collected, and online iterative optimization is performed to obtain a target energy management scheme.

[0006] In a second aspect, the present invention provides an energy management and recovery system for a cold chain supply chain, the energy management and recovery system for a cold chain supply chain comprising: The acquisition module is used to collect and preprocess the operating parameters and environmental parameters of the cold chain system to obtain multi-dimensional feature data; A calculation module, used for performing feature extraction and local density distribution calculation on the multi-dimensional feature data to obtain energy consumption pattern clustering data; A construction module, used to construct a device operation knowledge graph based on the energy consumption pattern clustering data; An analysis module, used for performing abnormal feature analysis on the equipment operation knowledge graph and the energy consumption pattern clustering data through a hybrid neural network model to obtain abnormal feature data of equipment energy consumption; A generating module, used for generating an initial energy recovery plan according to the abnormal energy consumption characteristic data of the equipment; The optimization module is used to apply the initial energy recovery plan to the cold chain system, collect operation efficiency data and load matching data, and perform online iterative optimization to obtain a target energy management plan.

[0007] The third aspect of the present invention provides an energy management and recovery device for a cold chain supply chain, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the energy management and recovery device for the cold chain supply chain executes the above-mentioned energy management and recovery method for the cold chain supply chain.

[0008] A fourth aspect of the present invention provides a computer-readable storage medium, in which instructions are stored, which, when executed on a computer, enable the computer to execute the above-mentioned energy management and recovery method for the cold chain supply chain.

[0009] In the technical solution provided by the present invention, through a multi-level data collection and preprocessing mechanism, comprehensive perception of the operating parameters and environmental parameters of the cold chain system is achieved, providing a high-quality data foundation required for energy management; a clustering analysis method based on energy consumption distance is adopted to accurately identify the energy consumption pattern characteristics of the system, providing data support for the formulation of energy recovery strategies; combined with knowledge graph technology, a time series correlation model between devices is constructed, and the energy transfer law and equipment state evolution characteristics in the system are deeply explored; through the feature fusion mechanism of the hybrid neural network model, accurate identification of abnormal energy consumption states of equipment is achieved, providing a reliable basis for the generation of energy recovery schemes; based on multi-time scale operation efficiency evaluation and load matching analysis, a hierarchical energy management optimization mechanism is established to ensure the stability and efficiency of system operation; through a closed-loop online iterative optimization strategy, dynamic adjustment and continuous optimization of the energy recovery scheme are achieved, significantly improving the energy utilization efficiency of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0011] Figure 1 A schematic diagram of a process for an energy management and recovery method for a cold chain supply chain provided in an embodiment of the present application; Figure 2 A schematic block diagram of the structure of the energy management and recovery system for the cold chain supply chain provided in an embodiment of the present application; Figure 3 A schematic block diagram of the structure of the energy management and recovery equipment for the cold chain supply chain provided in an embodiment of the present application. DETAILED DESCRIPTION

[0012] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0013] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change based on actual conditions.

[0014] It should also be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0015] It should be further understood that the term “and / or” used in the specification and appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0016] In conjunction with the accompanying drawings, some embodiments of the present application are described in detail below. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0017] See also Figure 1 , Figure 1 A schematic diagram of a process flow of an energy management and recovery method for a cold chain supply chain provided in an embodiment of the present application, such as Figure 1 As shown, the energy management and recovery method for the cold chain supply chain provided in the embodiment of the present application includes steps S100 to S600.

[0018] Step S100: collecting and preprocessing the operation parameters and environmental parameters of the cold chain system to obtain multi-dimensional feature data; It is understandable that the execution subject of the present invention may be an energy management and recovery system of a cold chain supply chain, or a terminal or a server, which is not specifically limited here. The embodiment of the present invention is described by taking a server as the execution subject as an example.

[0019] Specifically, data is collected from the compressor unit to obtain key indicators such as the pressure of the refrigerant, the temperature of the refrigerant, the current of the motor, and the vibration frequency. After the data collection is completed, the data are normalized to ensure that data of different dimensions are compared under the same standard. After normalization, the operating characteristic data of the compressor unit is obtained, which reflects the real-time operating status of the compressor unit and provides basic data support for subsequent energy management. Data collection is performed on the operating status of the condenser. The collected parameters include key indicators such as heat exchange pressure, condensing temperature, fan speed, and load power. After the data collection is completed, the data are standardized to ensure the consistency and reliability of the data, and the operating characteristic data of the condenser is formed to reflect the important role of the condenser in the cold chain system and its operating efficiency. At the same time, data is collected on the operating status of the evaporator, including parameters such as evaporation pressure, evaporation temperature, defrost status, and refrigeration capacity. Through unified dimension processing, the consistency of different types of data is ensured during the analysis process, and the operating characteristic data of the evaporator is obtained. The operating characteristic data of the compressor unit, condenser, and evaporator are aligned in time series to ensure that the operating data of different devices are effectively compared and analyzed in the time dimension, forming a complete equipment operating characteristic matrix. Data collection is performed on the temperature, humidity and door status in the cold storage. In order to improve the smoothness of the data, the sliding window method is used to smooth the collected environmental parameters to obtain environmental parameter characteristic data. This process can help understand the energy consumption performance of the cold chain system under different environmental conditions. In terms of data accuracy, the outliers in the equipment operation characteristic matrix are corrected by local linear regression to effectively eliminate the noise and interference in the data and obtain more reliable first operation characteristic data. For the missing values ​​in the first operation characteristic data, the time series interpolation method is used to ensure the integrity and continuity of the data and obtain the processed second operation characteristic data. The second operation characteristic data and the environmental parameter characteristic data are subjected to feature fusion and dimensionality reduction processing to form multidimensional characteristic data. Data from different sources are integrated so that the final data set contains both information on the equipment operation status and environmental impact. At the same time, dimensionality reduction processing is to reduce the complexity of the data and improve the efficiency of subsequent analysis, so that useful information can still be quickly obtained when the amount of data is large.

[0020] Step S200, extracting features and calculating local density distribution of multi-dimensional feature data to obtain energy consumption pattern clustering data; Specifically, the multidimensional feature data is grouped according to the equipment operation characteristics, environmental characteristics and energy consumption characteristics. The mean and standard deviation of each group of features are calculated to obtain feature statistical parameters to reflect the distribution and change trend of different features. The energy consumption feature normalization matrix is ​​constructed according to the feature statistical parameters to perform standardized transformation on the multidimensional feature data, ensure that different features are compared under the same standard, and obtain standardized feature data. Based on the standardized feature data, the feature weighting matrix is ​​constructed by setting the equipment operation feature weight coefficient, the environmental feature weight coefficient and the energy consumption feature weight coefficient to obtain weighted feature data. The focus on important features is enhanced to ensure that the energy consumption analysis can reflect the actual situation. The sample distance matrix is ​​obtained by using the Euclidean distance between any two data samples in the weighted feature data and the feature weighting matrix. In order to improve the accuracy of clustering, a search radius threshold is set, and the number of neighboring samples of each sample point within the range is counted to identify the density distribution of samples in the local area and obtain local density distribution data. According to the local density distribution data, a set of sample points with a larger density value is found for each sample point, and the distance to the nearest sample point in the set is calculated to generate density distance data. Combining the product of local density value and density distance, we can effectively identify density cluster points and obtain cluster center data, which reflects the main characteristics of energy consumption patterns. Based on cluster center data and sample distance matrix, we assign the remaining sample points to the nearest cluster center to ensure that each sample point is reasonably classified into the energy consumption pattern that best matches its characteristics, and obtain the final energy consumption pattern cluster data.

[0021] Step S300, constructing a device operation knowledge graph according to energy consumption pattern clustering data; Specifically, the equipment information, energy consumption data and operating parameters in the energy consumption pattern clustering data are classified and coded. Different categories of information are effectively distinguished and standardized in coded form to facilitate subsequent graph construction. A node attribute table is constructed, and the nodes represent the equipment in the system and its operating characteristics to form basic node data. Based on the basic node data, an equipment composition relationship diagram is constructed. The physical composition relationship and operating parameters of the compressor unit, condenser and evaporator are converted into physical connection relationships between nodes to obtain equipment relationship data. The equipment relationship data reflects the hardware connection between each device and clarifies how the operating parameters of different devices affect each other. The energy consumption transfer path is established using the equipment relationship data, and the energy flow process in the cold chain system is converted into a directed edge relationship to obtain energy consumption relationship data. The energy consumption relationship data describes the flow of energy between different devices, so that the energy transfer process can be modeled in the form of a graph, providing an overall perspective on the energy consumption of the cold chain system and a better understanding of the interaction between devices and the energy transfer path. According to the energy consumption relationship data, an equipment state transfer link is established to convert the dynamic changes in the equipment operation state into a time-series state sequence to obtain state migration data. By recording the state changes of the equipment in different time periods, it is possible to analyze the performance of the equipment under different operating conditions and how these changes affect the energy consumption of the entire system. Multimodal information integration is performed on the state transition data, and physical quantities such as temperature, power, and pressure are mapped to the dynamic attributes of the corresponding nodes to obtain node feature data. Association rule mining is performed based on the node feature data to build a temporal dependency matrix between nodes to obtain association rule data, revealing the dependency between different device nodes, especially their interaction in time series, thereby providing a basis for the coordinated operation of devices. The association rule data is converted into a graph data structure, the temporal edge weights between nodes are constructed, and the connection relationship between nodes is sparsely processed to obtain graph structure data. Sparse processing reduces unnecessary connections between nodes and reduces computational complexity, making the structure of the knowledge graph clearer, thereby improving the efficiency of graph analysis. Node embedding and edge embedding calculations are performed on the graph structure data, and the temporal features of nodes and edges in the graph are vectorized. Through node embedding and edge embedding, the characteristics of equipment operation are mapped to a high-dimensional vector space. These vector representations retain the relationship between equipment and the timing characteristics of energy transfer, and enable complex equipment interactions to be analyzed and modeled through simple mathematical operations. This forms an equipment operation knowledge graph. This knowledge graph contains the physical connection relationship, energy consumption transfer path, and state transfer link of the equipment, and integrates multi-modal operation feature data. Through the data representation of the graph structure, the timing characteristics and association rules between equipment are integrated. The equipment operation knowledge graph can intuitively reflect the energy transfer and collaborative operation relationship between equipment in the cold chain system.

[0022] Step S400: performing abnormal feature analysis on the equipment operation knowledge graph and energy consumption pattern clustering data through a hybrid neural network model to obtain abnormal feature data of equipment energy consumption; Specifically, for the equipment operation knowledge graph, node-level and edge-level graph convolution layers are constructed respectively. The node-level graph convolution layer uses three layers of convolution operations, and the sizes of the convolution kernels are 32, 64, and 128, respectively, to gradually extract higher-level graph structure features. The edge-level graph convolution layer contains two layers of convolution operations, and the sizes of the convolution kernels are 16 and 32, respectively, to capture the edge information between nodes and obtain more comprehensive graph structure features. These graph structure features contain representations of equipment nodes in the cold chain system and the complex relationships between them. The graph structure features are input into the relationship perception layer, which includes the equipment composition relationship sublayer, the energy consumption transmission relationship sublayer, and the state transition relationship sublayer. Each sublayer is set with 64 feature channels, and the information from different relationships is fused through the message passing mechanism to obtain a comprehensive graph feature vector. At the same time, the energy consumption pattern clustering data is constructed to construct a multi-branch feature extraction network. The network includes compressor unit feature branch, condenser feature branch and evaporator feature branch. Each branch contains two fully connected layers with 128 and 64 neurons respectively. Important features are extracted from different characteristics of the equipment to obtain equipment feature vectors to describe the operating status of each equipment. The equipment feature vector is input into the time series feature extraction layer. This layer adopts a gated recurrent unit structure, including input gate, reset gate and update gate to effectively capture the time series change characteristics of equipment features. The hidden state dimension is set to 128 and the time series window length is 24, which means that the operating data within 24 time steps is modeled to obtain the time series feature vector, revealing the time dependence and trend change of the equipment during operation. The graph feature vector and the time series feature vector are input into the cross attention module. This module sets 8 attention heads, each with a dimension of 32, and adaptively fuses by calculating the correlation weights between features. The role of the cross attention mechanism is to dynamically pay attention to the importance between graph features and time series features to capture the interactive information between different modal data and obtain multimodal fusion features to represent the complex state and energy consumption pattern of equipment in the cold chain system. The multimodal fusion features are input into the feature enhancement network for processing. The feature enhancement network contains three residual blocks, each of which consists of two convolutional layers and a short-circuit connection, and a batch normalization layer and a ReLU activation function are set between the residual blocks. The residual block is designed to solve the gradient vanishing problem in deep networks while retaining the useful information of the input features. Through the operation of the residual block, an enhanced feature vector is obtained. The enhanced feature vector is input into the abnormal pattern recognition layer, which contains an energy consumption abnormality subnetwork, a temperature abnormality subnetwork, and a pressure abnormality subnetwork. Each subnetwork consists of three fully connected layers, and the number of neurons is 256, 128, and 64, respectively. Each subnetwork is dedicated to identifying a specific type of abnormal pattern. The energy consumption abnormality subnetwork is used to detect whether the energy consumption of the device is abnormal, the temperature abnormality subnetwork is used to identify abnormal temperature fluctuations, and the pressure abnormality subnetwork focuses on the pressure state of the device.These sub-networks enhance the feature vectors through comprehensive analysis to obtain the abnormal pattern vectors. Comprehensive analysis is performed on the abnormal pattern vectors. The weight coefficients of energy consumption anomaly, temperature anomaly, and pressure anomaly are set to 0.4, 0.3, and 0.3 respectively. Through the weighted combination method, the importance of different anomaly types is comprehensively considered, and finally the abnormal energy consumption feature data of the equipment is obtained.

[0023] Step S500: generating an initial energy recovery plan according to the abnormal energy consumption characteristic data of the equipment; Specifically, a recovery priority matrix is ​​constructed for the abnormal energy consumption feature data of the equipment. The abnormal degree is graded based on the weight of 0.4 for energy consumption abnormality, 0.3 for temperature abnormality, and 0.3 for pressure abnormality to obtain abnormal priority data. By identifying the priorities corresponding to different types of abnormalities, the energy consumption abnormalities with the greatest impact are solved first in the process of energy recovery, effectively improving the recovery efficiency. An energy recovery type mapping table is established based on the abnormal priority data, and different abnormal characteristics are matched with corresponding energy recovery types. In the mapping process, a corresponding relationship is established between energy consumption abnormalities and three recovery modes, namely, compressor waste heat recovery, condensation heat recovery, and system energy balance, to obtain recovery type data. The recovery type data is feature expanded, and a feature enhancement matrix is ​​constructed by introducing historical recovery efficiency indicators and equipment operating status indicators to obtain recovery feature data. Combined with the historical performance and current operating status of the equipment, the descriptive ability of the recovery type data is enhanced, so that the generation of the recovery strategy is more in line with the actual operating conditions and historical efficiency performance of the equipment. The recovery feature data is input into the input layer of the Softmax regression classifier. After being processed by two hidden layers, each layer has 128 neurons, and the BatchNorm and ReLU activation functions are configured to obtain the classification feature data. BatchNorm speeds up the training speed of the model and stabilizes the training process, while the ReLU activation function introduces nonlinear factors, enabling the classifier to model complex feature relationships and obtain classification feature data. Based on the classification feature data, the probability distribution of the recovery strategy is generated. The compressor waste heat recovery parameters, condensation heat recovery parameters, and system energy balance parameters are scored probabilistically to obtain strategy probability data. The generation process of strategy probability data reflects the applicability and probability distribution of different recovery strategies under the current energy consumption abnormality characteristics. Based on the strategy probability data, a parameter configuration matrix is ​​constructed, which contains the value ranges of the compressor speed, condensation temperature, and evaporation temperature to obtain parameter configuration data. The configuration ranges of these parameters provide feasible operating parameters for the specific implementation of energy recovery to ensure that the recovery process is carried out under reasonable conditions. Operation constraint analysis is performed on the parameter configuration data. Based on the operating boundary conditions of the equipment and the stability requirements of the system, the corresponding parameter adjustment step is set to obtain the operating parameter data. The purpose of setting a suitable parameter adjustment step is to ensure the smooth operation of the equipment during the energy recovery process and avoid system instability caused by drastic changes in parameters. On this basis, according to the timing of system response and the link of energy transfer, the execution of operating parameters is organized in a reasonable order, and finally a control instruction sequence is constructed to obtain the initial energy recovery plan.

[0024] Step S600: Apply the initial energy recovery plan to the cold chain system, collect operation efficiency data and load matching data, and perform online iterative optimization to obtain a target energy management plan.

[0025] Specifically, the operating parameter data in the initial energy recovery scheme is sent to the control system of the compressor unit, condenser and evaporator. In this way, the key equipment of the cold chain system is controlled in real time, and the compressor waste heat recovery data, condensation heat recovery data and system energy balance data are collected to obtain the first operating data. The first operating data is collected in segments to facilitate the evaluation of the operation of the equipment at different time scales. The segmented collection is carried out according to three time scales of 5 minutes, 15 minutes and 30 minutes, and the dynamic change process of the equipment operation efficiency parameters and load matching parameters is recorded to obtain the second operating data. The segmented collection method can reflect the operation status of the cold chain system in the short, medium and long term. Based on the second operating data, the short-term, medium and long-term operating efficiency indicators are calculated respectively, including the compressor energy efficiency ratio, the condenser heat exchange efficiency and the evaporator refrigeration coefficient, so as to obtain efficiency evaluation data. These indicators can measure the energy efficiency performance of the equipment at different time scales and provide detailed information on the energy utilization rate of the equipment. The compressor energy efficiency ratio reflects the relationship between the energy consumption and refrigeration output of the compressor, the condenser heat exchange efficiency reflects its heat exchange performance, and the evaporator refrigeration coefficient is used to evaluate the effectiveness of its refrigeration capacity. Multi-scale load matching analysis is performed on the efficiency evaluation data. By calculating the deviations of the real-time load, hourly load and daily load from the set value and performing dynamic load balance calculation, the load matching data is obtained to ensure that the load of the cold chain system is always consistent with the set value, so as to maximize the energy efficiency of the system under various operating conditions. Through load matching analysis, the deviation between the load and the actual operating demand is identified, and corresponding adjustment measures are taken to maintain the stable operation and efficient energy utilization of the system. The efficiency evaluation data and load matching data are input into the adaptive optimization iterator, and the energy recovery parameters are dynamically adjusted by the stochastic gradient descent method with momentum term to obtain the first optimization parameter data. The use of stochastic gradient descent method can effectively reduce the local minimum problem encountered in the optimization process, while the momentum term accelerates the convergence process, so as to find the global optimal solution faster. Through the adaptive optimization iterator, the energy recovery parameters can be continuously adjusted to make the equipment operation more optimal. The first optimization parameter data is subjected to hierarchical operation constraint verification. According to the safe operation boundary of the equipment, the system stability requirements and the energy balance constraints, the optimization parameters are corrected to ensure that the optimized parameters operate within the safety range of the equipment and meet the stability requirements of the system at the same time. Through multiple checks, the second optimization parameter data is obtained to ensure that the optimization process will not have a negative impact on the safety of the equipment and the stability of the system. According to the second optimization parameter data, a hierarchical control instruction sequence is constructed. According to the different response requirements of the cold chain system, the execution strategy is reconstructed according to the timing characteristics of the fast response layer, the mid-term adjustment layer, and the slow optimization layer to obtain hierarchical control data.The hierarchical control method enables the system to make corresponding adjustments according to different time scale requirements. The fast response layer is used to deal with instantaneous changes, the medium-term adjustment layer is used to adapt to fluctuations in a longer time range, and the slow optimization layer focuses on improving the long-term energy efficiency of the system. The hierarchical control data and the initial energy recovery plan are optimized through multi-objective fusion, and a closed-loop control process with a prediction-feedback-compensation mechanism is constructed to finally obtain the target energy management plan. Through closed-loop control, energy utilization is continuously optimized in the process of continuous prediction, feedback and compensation, ensuring that the system can maintain optimal energy efficiency under different operating conditions.

[0026] In the embodiment of the present invention, through a multi-level data collection and preprocessing mechanism, comprehensive perception of the operating parameters and environmental parameters of the cold chain system is achieved, providing a high-quality data foundation required for energy management; a clustering analysis method based on energy consumption distance is adopted to accurately identify the energy consumption pattern characteristics of the system, providing data support for the formulation of energy recovery strategies; combined with knowledge graph technology, a time series correlation model between devices is constructed, and the energy transfer law and equipment state evolution characteristics in the system are deeply explored; through the feature fusion mechanism of the hybrid neural network model, accurate identification of abnormal energy consumption states of equipment is achieved, providing a reliable basis for the generation of energy recovery plans; based on multi-time scale operation efficiency evaluation and load matching analysis, a hierarchical energy management optimization mechanism is established to ensure the stability and efficiency of system operation; through a closed-loop online iterative optimization strategy, dynamic adjustment and continuous optimization of the energy recovery plan are achieved, significantly improving the energy utilization efficiency of the system.

[0027] In a specific embodiment, the process of executing step S100 may specifically include the following steps: The refrigerant pressure, refrigerant temperature, motor current and vibration frequency of the compressor unit are collected and normalized to obtain the operating characteristic data of the compressor unit; the heat exchange pressure, condensing temperature, fan speed and load power of the condenser are collected and normalized to obtain the operating characteristic data of the condenser; the evaporation pressure, evaporation temperature, defrost state and cooling capacity of the evaporator are collected and dimensionally unified to obtain the operating characteristic data of the evaporator; Performing time-series alignment processing on the compressor unit operation characteristic data, the condenser operation characteristic data and the evaporator operation characteristic data to obtain the equipment operation characteristic matrix; Collect data on the temperature and humidity in the cold storage and the door status, and use the sliding window method to smooth the data to obtain the characteristic data of environmental parameters; Performing local linear regression correction on abnormal values ​​in the equipment operation characteristic matrix to obtain first operation characteristic data, and performing time series interpolation to complete missing values ​​in the first operation characteristic data to obtain second operation characteristic data; The second operation characteristic data and the environmental parameter characteristic data are subjected to feature fusion and dimensionality reduction processing to obtain multi-dimensional characteristic data.

[0028] Specifically, the operating parameters of key equipment in the cold chain system are collected and preprocessed, including the operating data of the compressor unit, condenser and evaporator. In the process of collecting data from the compressor unit, key parameters such as refrigerant pressure, refrigerant temperature, motor current and vibration frequency are collected. In order to ensure the consistency of the data and the effectiveness of the model, these data are normalized to convert the operating parameters of the compressor unit to a unified range, usually [0,1]. Through normalization, the operating characteristic data of the compressor unit is obtained. The operating data of the condenser is collected, including heat exchange pressure, condensing temperature, fan speed and load power. These data are standardized to convert data of different dimensions into a standard normal distribution, so as to maintain the importance and comparability of each feature in subsequent analysis. The standardized calculation formula is: ; in, Represents the original data, is the mean of the data, is the standard deviation, is the standardized value. Through the standardization process, the operating characteristic data of the condenser is obtained, which reflects the operating status of the condenser in different time periods. At the same time, the operating status of the evaporator is detected, and the collected parameters include evaporation pressure, evaporation temperature, defrost status and cooling capacity. For the data of the evaporator, due to the inconsistent dimensions of different features, such as pressure in Pascals, temperature in Celsius, and cooling capacity in Watts, these data are dimensionalized to ensure effective comparison and modeling between the features. Through the process of dimensional unification, the operating characteristic data of the evaporator is obtained. The operating characteristic data of the compressor unit, condenser and evaporator are time-series aligned to construct the equipment operating characteristic matrix, which contains the time series operating data of the equipment. Data collection is performed on the temperature, humidity and door status in the cold storage, and the data is smoothed by the sliding window method to obtain the environmental parameter characteristic data. The calculation of the sliding window method is expressed as: in, For time point The smoothed value on is the original data, is the size of the sliding window. The sliding window method is used to eliminate random fluctuations in environmental data, making the data smoother and more stable, and better reflecting the real change trend of the environment. Correct the outliers in the equipment operation feature matrix. The local linear regression method is used to correct the outliers. Local linear regression uses adjacent data points to fit a linear regression line to correct the abnormal data and obtain the first operation feature data. The missing values ​​in the first operation feature data are interpolated and supplemented in time series to ensure the continuity and integrity of the data. The interpolation method uses linear interpolation or spline interpolation to obtain the second operation feature data. The second operation feature data is subjected to feature fusion and dimensionality reduction processing with the environmental parameter feature data to obtain multidimensional feature data. The equipment operation characteristics and environmental characteristics are combined to fully reflect the operating status and environmental impact of the cold chain system. The dimensionality reduction processing uses the principal component analysis method to project the high-dimensional data into a lower-dimensional subspace in order to reduce the computational complexity while maintaining the main information of the data to obtain multidimensional feature data.

[0029] In a specific embodiment, the process of executing step S200 may specifically include the following steps: The multi-dimensional feature data is grouped according to the equipment operation characteristics, environmental characteristics and energy consumption characteristics, and the mean and standard deviation of each group of characteristics are calculated to obtain the feature statistical parameters; An energy consumption characteristic normalization matrix is ​​constructed according to characteristic statistical parameters, and the multi-dimensional characteristic data is standardized to obtain standardized characteristic data. Based on the standardized characteristic data, a characteristic weighting matrix is ​​constructed through the equipment operation characteristic weight coefficient, the environmental characteristic weight coefficient and the energy consumption characteristic weight coefficient to obtain weighted characteristic data; The Euclidean distance between any two data samples in the weighted feature data and the feature weighting matrix are used to perform weighted distance calculation to obtain a sample distance matrix, and a search radius threshold is set for the sample distance matrix. The number of neighboring samples of each sample point within the search radius threshold is counted to obtain local density distribution data. According to the local density distribution data, for each sample point, find a set of sample points with a larger density value, and calculate the distance to the nearest sample point in the set to obtain density distance data; The local density distribution data and the density distance data are jointly analyzed, and the density clustering points are determined by the product of the local density value and the density distance to obtain the cluster center data. Based on the cluster center data and the sample distance matrix, the remaining sample points are assigned to the nearest cluster center to obtain the energy consumption pattern clustering data.

[0030] Specifically, the multidimensional feature data is grouped according to the equipment operation characteristics, environmental characteristics and energy consumption characteristics. After dividing the equipment operation characteristics (such as compressor speed, evaporator pressure, etc.), environmental characteristics (such as cold storage temperature, humidity, etc.) and energy consumption characteristics (such as power consumption), the mean and standard deviation of each group of characteristics are calculated to obtain the feature statistical parameters. According to the feature statistical parameters, the energy consumption feature normalization matrix is ​​constructed to perform a standardized transformation on the multidimensional feature data to obtain standardized feature data. The feature data is linearly transformed so that its mean is 0 and the standard deviation is 1, which helps to reduce the influence between different feature dimensions and ensure the stability of the subsequent algorithm. Through the standardized transformation, all feature data are mapped to the same scale, eliminating the influence of different features due to dimensional differences, and obtaining standardized feature data. Based on the standardized feature data, feature weighting processing is performed. By setting the equipment operation feature weight coefficient, environmental feature weight coefficient and energy consumption feature weight coefficient, a feature weighting matrix is ​​constructed to obtain weighted feature data, highlighting the importance of different features in energy consumption management. The feature weighting matrix is ​​expressed as: ; in, is the weighted feature data, To standardize feature data, For the The weight coefficient of the group feature. The Euclidean distance between any two data samples in the weighted feature data is used in combination with the feature weighting matrix to perform weighted distance calculation to obtain the sample distance matrix. The calculation formula of the Euclidean distance is: ; in, Representation sample and samples The weighted Euclidean distance between and The samples are and samples In the The weighted value of each trait, is the number of features. By calculating the sample distance, a sample distance matrix is ​​obtained to characterize the similarity between samples. After obtaining the sample distance matrix, a search radius threshold is set to count the number of neighboring samples of each sample point within the search radius to obtain local density distribution data. The size of the local density reflects the degree of clustering of samples in space. The smaller the search radius, the more clustered the samples are, the more they can be identified. According to the local density distribution data, for each sample point, a set of sample points with a larger density value is searched, and the distance to the nearest sample point in the set is calculated to obtain density distance data. The local density distribution data and the density distance data are jointly analyzed, and the density clustering point is determined by the product of the local density value and the density distance to obtain the cluster center data. The formula is expressed as: ; in, Represents sample points The aggregate value of For sample points The local density value of For sample points The distance between the sample points with a larger density value. By calculating the clustering value of each sample point, the point with the highest degree of clustering is found as the center point of the cluster. Based on the cluster center data and the sample distance matrix, the remaining sample points are assigned to the cluster center with the nearest distance to obtain the energy consumption mode clustering data, which reflects the inherent relationship between the equipment operating status, environmental conditions and energy consumption mode in the cold chain system.

[0031] In a specific embodiment, the process of executing step S300 may specifically include the following steps: The equipment information, energy consumption data and operating parameters in the energy consumption pattern clustering data are classified and coded, a node attribute table is constructed, and basic node data is obtained; Build an equipment composition relationship diagram based on basic node data, convert the composition relationship and operating parameters of the compressor unit, condenser, and evaporator into physical connection relationships between nodes, and obtain equipment relationship data; Use equipment relationship data to establish energy consumption transfer paths, transform the energy flow process in the cold chain system into directed edge relationships, and obtain energy consumption relationship data; Establish a device state transfer link based on energy consumption relationship data, convert the dynamic changes of the device operation state into a time-series state sequence, and obtain state migration data; Integrate multimodal information on state transition data, map temperature, power, and pressure to the dynamic attributes of corresponding nodes, obtain node feature data, and mine association rules based on node feature data to build a temporal dependency matrix between nodes to obtain association rule data; The association rule data is converted into a graph data structure, the temporal edge weights between nodes are constructed, and the connection relationships between nodes are sparsely processed to obtain graph structure data. Node embedding and edge embedding calculations are performed on the graph structure data, and the temporal features of the nodes and edges in the graph are vectorized to obtain the equipment operation knowledge graph.

[0032] Specifically, the equipment information, energy consumption data and operating parameters in the energy consumption pattern clustering data are classified and coded. Classify various types of information, such as independently classifying the equipment information of compressor units, condensers, evaporators, etc., classify the energy consumption data of each device (such as energy consumption rate, power utilization rate) into one group, and classify the operating parameters (such as temperature, pressure, wind speed, etc.) into another group. Standardize and encode each type of feature so that it can be uniformly managed and processed when constructing the graph later, build a node attribute table, and obtain basic node data. Build an equipment composition relationship diagram based on the basic node data, and convert the physical composition relationship and operating parameters of equipment such as compressor units, condensers, evaporators, etc. into physical connection relationships between nodes. Each device is regarded as a node, and the connection relationship between nodes is represented by an edge. For example, the connection between the compressor and the condenser represents the heat exchange process of the compressor refrigerant through the condenser. These physical connection relationships are represented in the form of an adjacency matrix. The adjacency matrix A is defined as follows: ; in, Indicates the device and equipment If and If there is a physical connection between is 1; otherwise it is 0. In this way, the equipment relationship data is obtained, and a complete equipment composition relationship diagram is constructed to reflect the physical connection and energy transfer path between each device. Use equipment relationship data to establish the energy consumption transfer path in the cold chain system. The energy transfer between devices is regarded as a directed relationship, such as the energy flow process from the compressor to the condenser and then to the evaporator. By establishing a directed edge relationship, the energy consumption transfer process is mapped to a graph structure to obtain energy consumption relationship data. The edges in the directed graph are represented by a directed adjacency matrix. The elements in the matrix represent whether there is a connection between the nodes and indicate the direction of energy flow. Based on the energy consumption relationship data, a state transfer link for the equipment is established, and the dynamic changes in the equipment operation state are converted into a time-series state sequence to obtain state migration data. The state transfer link describes the changes in the state of the equipment during operation, such as the process from the start-up of the compressor, load change to stop, and these changes are represented by a state sequence. State migration matrix Indicates the device At the moment The state is transferred to the device The state is defined by the following formula: ; in, Indicates the device At the moment Transfer to device exist The probability of time, and Respectively represent devices and equipment The state at each moment. These state transition information constitute the temporal mutual influence between devices. Multimodal information integration is performed on the state transition data, and the operating parameters of the device such as temperature, power, pressure, etc. are mapped to the dynamic attributes of the corresponding node to obtain node feature data. Through multimodal information integration, the state and characteristics of the node are described. Association rule mining is performed based on the node feature data to construct a temporal dependency matrix between nodes. Find the temporal association between devices, for example, the operating state of a certain device affects the energy consumption performance of another device. By mining this association, association rule data is obtained to reflect the dependency between devices. The association rule data is converted into a graph data structure to construct the temporal edge weights between nodes. For example, through the weight matrix Describe the timing dependency between nodes, the formula is as follows: ; in, Representation Node and the edge weight between node j, Representation Node and The strength of the association between It is a function that maps the strength of association to the weight of the edge. In order to reduce the computational complexity and improve the representation ability of the graph structure, the connection relationship between nodes is sparsely processed, the most important connections are retained and the connections with less influence are ignored, so as to obtain more concise graph structure data. Node embedding and edge embedding calculations are performed on the graph structure data. The nodes and edges in the graph are represented as vectors to facilitate subsequent processing by machine learning models. Node embedding uses an embedding method based on graph convolution, such as calculating the embedding vector of node features through a graph convolutional neural network. Edge embedding is a comprehensive calculation of the edge attributes and the node features of the connection to obtain the edge embedding vector. Through node embedding and edge embedding, the time series features in the graph are vectorized to obtain the knowledge graph of device operation.

[0033] In a specific embodiment, the process of executing step S400 may specifically include the following steps: The node-level and edge-level graph convolution layers are constructed for the device operation knowledge graph. The node-level graph convolution layer contains three layers of convolution operations, and the convolution kernel sizes are 32, 64, and 128 respectively. The edge-level graph convolution layer contains two layers of convolution operations, and the convolution kernel sizes are 16 and 32 respectively, to obtain the graph structure features. The graph structure features are input into the relationship perception layer, which includes the device composition relationship sublayer, energy consumption transmission relationship sublayer and state transfer relationship sublayer. Each sublayer is set with 64 feature channels. Multiple relationship information is integrated through the message transmission mechanism to obtain the graph feature vector. A multi-branch feature extraction network is constructed for the energy consumption pattern clustering data. The multi-branch feature extraction network includes a compressor unit feature branch, a condenser feature branch, and an evaporator feature branch. Each branch contains two fully connected layers with 128 and 64 neurons, respectively, to obtain the device feature vector. The device feature vector is input into the timing feature extraction layer. The timing feature extraction layer adopts a gated recurrent unit structure, including an input gate, a reset gate, and an update gate. The hidden state dimension is set to 128, and the timing window length is 24 to obtain a timing feature vector. The graph feature vector and the time series feature vector are input into the cross attention module. The cross attention module is set with 8 attention heads, each with a dimension of 32. The correlation weights between features are calculated for adaptive fusion to obtain multimodal fusion features. The multimodal fusion features are input into the feature enhancement network, which contains three residual blocks. Each residual block consists of two convolutional layers and a short-circuit connection. A batch normalization layer and a ReLU activation function are set between the residual blocks to obtain an enhanced feature vector. The enhanced feature vector is input into the abnormal pattern recognition layer. The abnormal pattern recognition layer includes the energy consumption abnormality sub-network, the temperature abnormality sub-network and the pressure abnormality sub-network. Each sub-network consists of three fully connected layers with 256, 128 and 64 neurons respectively, and the abnormal pattern vector is obtained. A comprehensive analysis is performed on the abnormal pattern vectors, and the weight coefficients of energy consumption anomaly, temperature anomaly, and pressure anomaly are set to 0.4, 0.3, and 0.3, respectively. The abnormal energy consumption characteristic data of the equipment is obtained through weighted combination.

[0034] Specifically, node-level and edge-level graph convolution layers are constructed for the equipment operation knowledge graph. The node-level graph convolution layer contains three layers of convolution operations, and the convolution kernel sizes are 32, 64, and 128, respectively, to gradually extract the deep features of the nodes in the graph. The edge-level graph convolution layer contains two layers of convolution operations, and the convolution kernel sizes are 16 and 32, respectively, to capture the relationship information of the edges between nodes. The node-level and edge-level graph convolution layer settings help to extract multi-level graph structure features from the equipment operation knowledge graph. Node-level convolution is used to extract the characteristics of the equipment nodes (such as compressors, condensers, evaporators, etc.) themselves, as well as their direct interaction characteristics with surrounding nodes; while edge-level convolution focuses on describing the physical and energy consumption transfer relationships between nodes. These operations can effectively extract the complex relationship information contained in the graph. The graph structure features after the convolution operation are input into the relationship perception layer, which includes the device composition relationship sublayer, the energy consumption transfer relationship sublayer, and the state transition relationship sublayer. Each sublayer is set with 64 feature channels to process the graph structure features from multiple dimensions. In the relationship perception layer, multiple relationship information is fused through the message passing mechanism. When updating the features, each node not only considers its own features, but also the information of neighboring nodes, so as to fully integrate the complex associations between devices in the composition, energy consumption transfer and state change process, and obtain the graph feature vector. The message passing process is expressed by the following formula: ; in, For Node In the The feature vector of the layer, Representation Node The set of neighbor nodes of For the The weight matrix of the layer, is the bias term, is the activation function. Through this mechanism, the composition relationship, energy consumption transfer and state transition information of the equipment are integrated to form a comprehensive graph feature vector. A multi-branch feature extraction network is constructed for the energy consumption pattern clustering data. The multi-branch feature extraction network includes a compressor unit feature branch, a condenser feature branch and an evaporator feature branch. Each branch contains two fully connected layers with 128 and 64 neurons respectively, which are used to extract the feature vector of the equipment. Through the design of the multi-branch network, the features of different equipment are processed separately, so that the uniqueness of the equipment is retained when extracting features, while ensuring the consistency between the features of different equipment. These feature vectors describe the performance of each device in the energy consumption mode and are an important representation of the device status. The device feature vector is input into the timing feature extraction layer. The timing feature extraction layer adopts a gated recurrent unit structure, which includes an input gate, a reset gate and an update gate. The hidden state dimension is set to 128 and the timing window length is 24. The gated recurrent unit can effectively capture the time dependency of device features. Especially when the device operating state has strong timing characteristics, the gated recurrent unit can remember important information and forget irrelevant historical states through the gating mechanism, obtain the time representation of device features, and finally output the time feature vector. The graph feature vector and the time feature vector are input into the cross attention module together. The cross attention module sets 8 attention heads, each with a dimension of 32, and realizes adaptive fusion by calculating the correlation weights between features. The purpose of the cross attention mechanism is to establish connections between different modal data (such as graph features and time features) so that the mutual correlation between features can be better expressed. By learning the attention weights, it is adaptively judged which features are more important for the energy consumption anomaly identification of the device, and multimodal fusion features are obtained. The calculation of the attention mechanism is expressed by the following formula: ; in, , , denote the query matrix, key matrix and value matrix respectively, Represents the dimension of the key vector. Through the attention mechanism, each feature vector is weighted and summed to make the multimodal fusion feature more representative. The multimodal fusion feature is input into the feature enhancement network. The feature enhancement network contains three residual blocks. Each residual block consists of two convolutional layers and a short-circuit connection. A batch normalization layer and a ReLU activation function are set between the residual blocks. The design of the residual block can solve the gradient vanishing problem in the deep network. Through the short-circuit connection, the input feature is directly passed to the output layer, thereby retaining the original feature information. The batch normalization layer accelerates the convergence process of the network, and the ReLU activation function introduces nonlinearity, making the feature representation capability richer. After being processed by the feature enhancement network, the enhanced feature vector is finally obtained. The enhanced feature vector is input into the abnormal pattern recognition layer. The abnormal pattern recognition layer contains the energy consumption abnormality subnetwork, the temperature abnormality subnetwork and the pressure abnormality subnetwork. Each subnetwork consists of three fully connected layers, and the number of neurons is 256, 128 and 64 respectively. These subnetworks are used to identify different types of abnormal patterns. For example, the energy consumption abnormality subnetwork identifies abnormalities in the energy consumption of the device, the temperature abnormality subnetwork is used to identify whether the fluctuation of the device temperature is abnormal, and the pressure abnormality subnetwork focuses on the pressure changes of the device. Each subnetwork outputs an abnormal pattern vector, which reflects the abnormal state of the device in different aspects. A comprehensive analysis is performed on the abnormal pattern vectors. By setting the weight coefficients of energy consumption abnormality, temperature abnormality, and pressure abnormality to 0.4, 0.3, and 0.3 respectively, different types of abnormalities are weighted combined to obtain the energy consumption abnormality feature data of the device. The formula for weighted combination is: ; in, Represents the final abnormal feature data, , , The characteristic data of energy consumption anomaly, temperature anomaly and pressure anomaly are respectively represented. Through this weighted method, a comprehensive analysis is performed according to the relative importance of different anomalies to obtain more accurate anomaly identification results.

[0035] In a specific embodiment, the process of executing step S500 may specifically include the following steps: A recycling priority matrix is ​​constructed for the abnormal energy consumption characteristic data of the equipment, and the abnormal degree is graded based on the energy consumption abnormality weight of 0.4, the temperature abnormality weight of 0.3, and the pressure abnormality weight of 0.3 to obtain the abnormal priority data; An energy recovery type mapping table is established based on the abnormal priority data, and a corresponding relationship is established between the abnormal characteristics and the three recovery modes of compressor waste heat recovery, condensing heat recovery and system energy balance to obtain the recovery type data; The recycling type data is expanded, and the feature enhancement matrix is ​​constructed by introducing historical recycling efficiency indicators and equipment operation status indicators to obtain recycling feature data; The recovered feature data is input into the input layer of the Softmax regression classifier, and processed by two hidden layers, each layer is set with 128 neurons and configured with BatchNorm and ReLU activation functions to obtain the classification feature data; Generate recovery strategy probability distribution according to classification feature data, perform probability scoring on compressor waste heat recovery parameters, condensation heat recovery parameters and system energy balance parameters to obtain strategy probability data, and build parameter configuration matrix based on strategy probability data, including the value range of compressor speed, condensation temperature and evaporation temperature to obtain parameter configuration data; Perform operation constraint analysis on parameter configuration data, set parameter adjustment step according to equipment operation boundary conditions and system stability requirements, obtain operation parameter data, organize execution order of operation parameter data according to system response timing and energy transfer link, construct control instruction sequence, and obtain initial energy recovery plan.

[0036] Specifically, a recycling priority matrix is ​​constructed for the abnormal energy consumption characteristic data of the equipment to determine the priority of different equipment in energy recovery. When constructing the recycling priority matrix, weight coefficients are set for the different degrees of impact of abnormal energy consumption, abnormal temperature, and abnormal pressure, where the weight of abnormal energy consumption is 0.4, the weight of abnormal temperature is 0.3, and the weight of abnormal pressure is 0.3. Through these weights, the abnormal energy consumption characteristic data of the equipment is weighted and summed according to the degree of each abnormality to obtain the abnormal priority data of each device. The specific calculation formula is as follows: ; in, Indicates the device The exception priority of , , Respectively represent devices Abnormal characteristic values ​​in terms of energy consumption, temperature and pressure. The abnormalities of each device are weighted by this formula to obtain a priority data matrix. A mapping table of energy recovery types is established based on the abnormal priority data. The abnormal characteristics are corresponding to the three recovery modes of compressor waste heat recovery, condensation heat recovery and system energy balance. For example, if the energy consumption abnormality of the device has a higher priority, the compressor waste heat recovery method is preferred to utilize the excess heat; if the temperature abnormality has a higher priority, energy recovery through condensation heat recovery is considered; and for pressure abnormalities, the system energy balance method is used for processing. Through this mapping method, each abnormal characteristic is associated with the most appropriate energy recovery mode to obtain recovery type data. Feature expansion of recovery type data. By introducing historical recovery efficiency indicators and equipment operating status indicators, a feature enhancement matrix is ​​constructed to obtain recovery feature data. The historical recovery efficiency indicator reflects the actual effect of energy recovery under similar operating conditions, while the equipment operating status indicator describes the current operating status of the equipment, such as temperature, pressure, etc. The introduction of these features enables the recovery feature data to better combine historical experience with the current status, providing more comprehensive information for energy recovery decisions. The recovered feature data is input into the input layer of the Softmax regression classifier and processed by two hidden layers. Each layer has 128 neurons and is configured with batch normalization (BatchNorm) and ReLU activation functions. The purpose of batch normalization is to speed up the training process and prevent the gradient from disappearing, while the ReLU activation function can introduce nonlinear factors and enhance the expressiveness of the model. After being processed by these two hidden layers, the classified feature data is obtained. Based on the classified feature data, the Softmax regression classifier outputs the probability distribution of each energy recovery mode, including the probability scores of the compressor waste heat recovery parameters, the condensing heat recovery parameters, and the system energy balance parameters. Assumptions , , The probability scores of the three energy recovery modes are respectively expressed by the following formulas: ; in, represents the score of the energy recovery mode, Represents the probability of the corresponding mode. These probability scores reflect the applicability of different recovery modes in the current situation. Based on the strategy probability data, a parameter configuration matrix is ​​constructed, which contains the value ranges of the compressor speed, condensing temperature and evaporating temperature, and specific parameter configuration data is obtained. For example, at what speed should the compressor run, how should the temperature of the condenser be adjusted, etc. The parameter configuration data is subjected to operation constraint analysis to ensure that all parameters are within the operating boundary conditions of the equipment and meet the stability requirements of the system. By setting a reasonable parameter adjustment step, the equipment is ensured to remain stable during the adjustment process to avoid fluctuations caused by too fast adjustment. For example, the speed of the compressor needs to be adjusted in steps of 100 revolutions per minute to ensure a smooth transition of the system. After constraint analysis, the final operation parameter data is obtained. According to the operation parameter data, the execution order is organized according to the timing of the system response and the energy transfer link, and the control instruction sequence is constructed to form an initial energy recovery plan. In the control instruction sequence, the energy transfer relationship between devices is considered. For example, the waste heat of the compressor is sent to the condenser for condensation. Therefore, it is necessary to coordinate the operation order of each device to maximize the utilization of energy. The generation of control instructions needs to take into account the current state of the equipment and predict the response of the system so that optimal control can be achieved at each adjustment step.

[0037] In a specific embodiment, the process of executing step S600 may specifically include the following steps: The operating parameter data in the initial energy recovery scheme is sent to the control systems of the compressor unit, the condenser and the evaporator, and the compressor waste heat recovery data, the condensation heat recovery data and the system energy balance data are collected to obtain the first operating data; The first operation data is collected in sections according to three time scales of 5 minutes, 15 minutes and 30 minutes, and the dynamic change process of the equipment operation efficiency parameter and the load matching parameter is recorded to obtain the second operation data; Calculate the short-term operation efficiency index, the medium-term operation efficiency index and the long-term operation efficiency index respectively according to the second operation data, including the compressor energy efficiency ratio, the condenser heat exchange efficiency and the evaporator refrigeration coefficient, to obtain efficiency evaluation data; Perform multi-scale load matching analysis on the efficiency evaluation data, calculate the deviation between the real-time load, hourly load and daily load and the set value, and perform load dynamic balance calculation to obtain load matching data; The efficiency evaluation data and the load matching data are input into the adaptive optimization iterator, and the energy recovery parameters are dynamically adjusted by the stochastic gradient descent method with momentum term to obtain the first optimization parameter data; Performing hierarchical operation constraint verification on the first optimized parameter data, modifying the parameters according to the equipment safe operation boundary, system stability requirements and energy balance constraints, and obtaining the second optimized parameter data; Constructing a hierarchical control instruction sequence according to the second optimization parameter data, reconstructing the execution strategy according to the timing characteristics of the fast response layer, the mid-term adjustment layer and the slow optimization layer, and obtaining hierarchical control data; The hierarchical control data and the initial energy recovery plan are optimized through multi-objective fusion, and a closed-loop control process with a prediction-feedback-compensation mechanism is constructed to obtain the target energy management plan.

[0038] Specifically, the operating parameter data in the initial energy recovery scheme is sent to each device, including the control system of the compressor unit, condenser and evaporator. The operating parameters such as the speed of the compressor, the condensing temperature of the condenser and the cooling temperature of the evaporator are set by the control system to make each device enter the desired operating state. During the operation of the equipment, the waste heat recovery data of the compressor, the condensing heat recovery data of the condenser and the energy balance data of the entire system are collected in real time to obtain the first operating data. The first operating data reflects the energy utilization status of each device in actual operation, and also includes the mutual influence and coupling relationship between the devices. The first operating data is collected in sections according to three time scales of 5 minutes, 15 minutes and 30 minutes. Through the multi-time scale collection method, the dynamic changes of the operating efficiency parameters and load matching parameters of the equipment in different time periods are recorded to obtain the second operating data, which reflects the performance of the equipment in different time scales of short-term, medium-term and long-term. Based on the second operating data, the short-term, medium-term and long-term operating efficiency indicators are calculated respectively, including the energy efficiency ratio (COP) of the compressor, the heat exchange efficiency of the condenser and the refrigeration coefficient of the evaporator. The calculation of these indicators is to quantify the performance of the equipment at different time scales and evaluate its energy efficiency. The calculation formula of compressor energy efficiency ratio is: ; in, Indicates the energy efficiency ratio of the compressor, The cooling capacity provided by the compressor, is the input power of the compressor. By calculating this indicator, the energy efficiency performance of the compressor under different load conditions is reflected. For the condenser, the calculation of its heat exchange efficiency is expressed by the following formula: ; in, represents the heat transfer efficiency of the condenser, is the heat transfer capacity of the condenser, is the heat exchange area, The average temperature difference during the heat exchange process. The higher the heat exchange efficiency, the better the performance of the condenser. For the evaporator's refrigeration coefficient, use the following formula to calculate: ; in, It represents the refrigeration coefficient of the evaporator. The cooling capacity provided to the evaporator, is the energy input to the evaporator. These efficiency indicators constitute the efficiency evaluation data, which reflect the operating performance of each device in the cold chain system. A multi-scale load matching analysis is performed on the efficiency evaluation data. The load matching analysis evaluates whether the equipment is operating within the appropriate load range by calculating the deviation between the real-time load, hourly load and daily load of the equipment at different time scales and the set value, and obtains the load matching data. For example, for the compressor, its load matching degree in each time period is calculated to determine whether it is operating near the optimal energy efficiency point. Load matching analysis also involves dynamic balance calculation of the load to ensure that the load of each device can be adjusted smoothly at different time scales. The efficiency evaluation data and load matching data are input into the adaptive optimization iterator, and the energy recovery parameters are dynamically adjusted by the stochastic gradient descent method with momentum term to obtain the first optimized parameter data. The optimization process of the stochastic gradient descent method is expressed as: ; in, Indicated in The parameters in the iterations, is the learning rate, is the gradient of the loss function with respect to the parameters, is the weight of the momentum term, Represents the accumulated momentum term. By introducing momentum, the convergence speed is accelerated, and the oscillations that occur during the optimization process are reduced, so that the optimal value of the energy recovery parameter can be found more quickly. The first optimized parameter data is subjected to hierarchical operation constraint verification to ensure that the optimized parameters are within the safe operation boundary of the equipment and meet the constraints of system stability and energy balance. The second optimized parameter data is obtained by verifying whether the speed change of the compressor is within its allowable safety range and whether the temperature adjustment of the condenser will not have too much impact on other parts of the system. If the parameters do not meet these constraints, corresponding corrections are required to ensure the safety of equipment operation. According to the second optimized parameter data, a hierarchical control instruction sequence is constructed. The hierarchical control instructions include a fast response layer, a medium-term adjustment layer, and a slow optimization layer. These different control levels are used to achieve the optimal adjustment of the system at different time scales. For example, the fast response layer is used to process the instantaneous load changes of the compressor to ensure that the equipment adapts to load fluctuations in a short time; the medium-term adjustment layer is used to adjust the temperature of the condenser to cope with changes in ambient temperature; and the slow optimization layer is used for long-term energy efficiency optimization of the system to ensure that the entire system reaches the optimal operating state on a longer time scale. Through hierarchical control, hierarchical control data is obtained. The hierarchical control data and the initial energy recovery plan are optimized through multi-objective fusion to construct a closed-loop control process with prediction, feedback and compensation mechanisms to obtain the target energy management plan. The closed-loop control process continuously predicts the energy consumption and state changes of the equipment, provides real-time feedback on the operation of the equipment, and adjusts the parameters of the equipment through the compensation mechanism to achieve continuous optimization of the system. For example, when the load changes, the compressor adjusts its speed in time through the feedback of energy efficiency change information to avoid excessive energy consumption; when the ambient temperature of the condenser rises, the heat exchange efficiency of the condenser is adjusted through the prediction model to maintain the energy balance of the system. The introduction of closed-loop control ensures that the system can maintain optimal energy efficiency under different operating conditions.

[0039] See also Figure 2 , Figure 2 A schematic block diagram of the structure of the energy management and recovery system 200 for the cold chain supply chain provided in the embodiment of the present application, as shown in FIG. Figure 2 As shown, the energy management and recovery system 200 for the cold chain supply chain includes: The acquisition module 210 is used to collect and preprocess the operation parameters and environmental parameters of the cold chain system to obtain multi-dimensional feature data; The calculation module 220 is used to perform feature extraction and local density distribution calculation on the multi-dimensional feature data to obtain energy consumption pattern clustering data; A construction module 230 is used to construct a device operation knowledge graph based on energy consumption pattern clustering data; An analysis module 240 is used to perform abnormal feature analysis on the equipment operation knowledge graph and energy consumption pattern clustering data through a hybrid neural network model to obtain abnormal feature data of equipment energy consumption; A generating module 250 is used to generate an initial energy recovery plan according to the abnormal energy consumption characteristic data of the equipment; The optimization module 260 is used to apply the initial energy recovery plan to the cold chain system, collect operating efficiency data and load matching data, and perform online iterative optimization to obtain a target energy management plan.

[0040] Through the coordinated cooperation of the above-mentioned components and through a multi-level data collection and preprocessing mechanism, a comprehensive perception of the operating parameters and environmental parameters of the cold chain system is achieved, providing a high-quality data foundation required for energy management; a clustering analysis method based on energy consumption distance is adopted to accurately identify the energy consumption pattern characteristics of the system, providing data support for the formulation of energy recovery strategies; combined with knowledge graph technology, a time series correlation model between devices is constructed, and the energy transfer law and equipment state evolution characteristics in the system are deeply explored; through the feature fusion mechanism of the hybrid neural network model, accurate identification of abnormal energy consumption states of equipment is achieved, providing a reliable basis for the generation of energy recovery plans; based on multi-time scale operation efficiency evaluation and load matching analysis, a hierarchical energy management optimization mechanism is established to ensure the stability and efficiency of system operation; through a closed-loop online iterative optimization strategy, dynamic adjustment and continuous optimization of the energy recovery plan are achieved, significantly improving the energy utilization efficiency of the system.

[0041] See also Figure 3 , Figure 3 A schematic block diagram of the structure of an energy management and recovery device 300 for a cold chain supply chain provided in an embodiment of the present application, wherein the energy management and recovery device 300 for a cold chain supply chain includes a processor 301 and a memory 302, wherein the processor 301 and the memory 302 are connected via a system bus 303, wherein the memory 302 may include a non-volatile storage medium and an internal memory.

[0042] The non-volatile storage medium can store a computer program. The computer program includes program instructions, and when the program instructions are executed by the processor 301, the processor 301 can execute any of the above-mentioned energy management and recovery methods for the cold chain supply chain.

[0043] The processor 301 is used to provide computing and control capabilities to support the operation of the energy management and recovery device 300 of the entire cold chain supply chain.

[0044] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor 301, the processor 301 can execute any of the above-mentioned energy management and recovery methods for the cold chain supply chain.

[0045] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the present application scheme, and does not constitute a limitation on the energy management and recovery equipment 300 of the cold chain supply chain involved in the present application scheme. The specific energy management and recovery equipment 300 of the cold chain supply chain may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0046] It should be understood that the processor 301 may be a central processing unit (CPU), and the processor 301 may also 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. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0047] It should be noted that technical personnel in the relevant field can clearly understand that, for the convenience and conciseness of description, the specific working process of the energy management and recovery equipment 300 of the cold chain supply chain described above can refer to the corresponding process of the aforementioned cold chain supply chain energy management and recovery method, and will not be repeated here.

[0048] An embodiment of the present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by one or more processors, the one or more processors implement the energy management and recovery method for the cold chain supply chain provided in the embodiment of the present application.

[0049] The computer-readable storage medium may be an internal storage unit of the energy management and recovery device 300 of the cold chain supply chain in the aforementioned embodiment, such as a hard disk or memory of the energy management and recovery device 300 of the cold chain supply chain. The computer-readable storage medium may also be an external storage device of the energy management and recovery device 300 of the cold chain supply chain, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc., equipped with the energy management and recovery device 300 of the cold chain supply chain.

[0050] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0051] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc., various media that can store program codes.

[0052] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for energy management and recovery in a cold chain supply chain, characterized in that: include: Collect and preprocess the operating parameters and environmental parameters of the cold chain system to obtain multi-dimensional feature data; Performing feature extraction and local density distribution calculation on the multi-dimensional feature data to obtain energy consumption pattern clustering data; Constructing a device operation knowledge graph based on the energy consumption pattern clustering data; Performing abnormal feature analysis on the equipment operation knowledge graph and the energy consumption pattern clustering data through a hybrid neural network model to obtain abnormal feature data of equipment energy consumption; Generate an initial energy recovery plan based on the abnormal energy consumption characteristic data of the equipment; The initial energy recovery scheme is applied to the cold chain system, operation efficiency data and load matching data are collected, and online iterative optimization is performed to obtain a target energy management scheme.

2. The energy management and recovery method for cold chain supply chain according to claim 1 is characterized in that: The operation parameters and environmental parameters of the cold chain system are collected and preprocessed to obtain multi-dimensional feature data, including: The refrigerant pressure, refrigerant temperature, motor current and vibration frequency of the compressor unit are collected and normalized to obtain the operating characteristic data of the compressor unit; the heat exchange pressure, condensing temperature, fan speed and load power of the condenser are collected and normalized to obtain the operating characteristic data of the condenser; the evaporation pressure, evaporation temperature, defrost state and cooling capacity of the evaporator are collected and dimensionally unified to obtain the operating characteristic data of the evaporator; Performing time sequence alignment processing on the compressor unit operation characteristic data, the condenser operation characteristic data, and the evaporator operation characteristic data to obtain an equipment operation characteristic matrix; Collect data on the temperature and humidity in the cold storage and the door status, and use the sliding window method to smooth the data to obtain the characteristic data of environmental parameters; Performing local linear regression correction on abnormal values ​​in the equipment operation characteristic matrix to obtain first operation characteristic data, and performing time series interpolation to complete missing values ​​in the first operation characteristic data to obtain second operation characteristic data; The second operation characteristic data and the environmental parameter characteristic data are subjected to feature fusion and dimensionality reduction processing to obtain multi-dimensional characteristic data.

3. The energy management and recovery method for cold chain supply chain according to claim 2 is characterized in that: The step of extracting features and calculating local density distribution of the multidimensional feature data to obtain energy consumption pattern clustering data includes: The multidimensional feature data are grouped according to equipment operation characteristics, environmental characteristics and energy consumption characteristics, and the mean and standard deviation of each group of characteristics are calculated to obtain feature statistical parameters; An energy consumption characteristic normalization matrix is ​​constructed according to the characteristic statistical parameters, the multidimensional characteristic data is standardized and transformed to obtain standardized characteristic data, and based on the standardized characteristic data, a characteristic weighting matrix is ​​constructed through the equipment operation characteristic weight coefficient, the environmental characteristic weight coefficient and the energy consumption characteristic weight coefficient to obtain weighted characteristic data; Using the Euclidean distance between any two data samples in the weighted feature data and the feature weighting matrix, weighted distance calculation is performed to obtain a sample distance matrix, and a search radius threshold is set for the sample distance matrix, and the number of neighboring samples of each sample point within the search radius threshold is counted to obtain local density distribution data; According to the local density distribution data, for each sample point, a sample point set with a larger density value is found, and the distance to the nearest sample point in the set is calculated to obtain density distance data; The local density distribution data and the density distance data are jointly analyzed, and the density clustering points are determined by the product of the local density value and the density distance to obtain the cluster center data. Based on the cluster center data and the sample distance matrix, the remaining sample points are assigned to the cluster center with the nearest distance to obtain the energy consumption pattern clustering data.

4. The energy management and recovery method for cold chain supply chain according to claim 3 is characterized in that: The step of constructing a device operation knowledge graph based on the energy consumption pattern clustering data includes: Classify and encode the equipment information, energy consumption data and operating parameters in the energy consumption pattern clustering data, construct a node attribute table, and obtain basic node data; Constructing a device composition relationship diagram based on the basic node data, converting the composition relationship and operating parameters of the compressor unit, condenser, and evaporator into a physical connection relationship between nodes to obtain device relationship data; The energy consumption transfer path is established by using the equipment relationship data, and the energy flow process in the cold chain system is converted into a directed edge relationship to obtain energy consumption relationship data; Establishing a device state transfer link according to the energy consumption relationship data, converting the dynamic change of the device operation state into a time-series state sequence, and obtaining state migration data; Perform multimodal information integration on the state transition data, map the temperature, power, and pressure to the dynamic attributes of the corresponding nodes, obtain node feature data, and perform association rule mining based on the node feature data to construct a temporal dependency matrix between nodes to obtain association rule data; The association rule data is converted into a graph data structure, the temporal edge weights between nodes are constructed, and the connection relationship between nodes is sparsely processed to obtain graph structure data. Node embedding and edge embedding calculations are performed on the graph structure data, and the temporal features of the nodes and edges in the graph are vectorized to obtain a device operation knowledge graph.

5. The energy management and recovery method for cold chain supply chain according to claim 4 is characterized in that: The abnormal feature analysis of the equipment operation knowledge graph and the energy consumption pattern clustering data is performed by using a hybrid neural network model to obtain abnormal feature data of equipment energy consumption, including: Running the knowledge graph on the device to construct node-level and edge-level graph convolution layers respectively, wherein the node-level graph convolution layer includes three layers of convolution operations, and the convolution kernel sizes are 32, 64, and 128 respectively, and the edge-level graph convolution layer includes two layers of convolution operations, and the convolution kernel sizes are 16 and 32 respectively, to obtain graph structure features; The graph structure features are input into the relationship perception layer, which includes the device composition relationship sublayer, the energy consumption transmission relationship sublayer and the state transfer relationship sublayer. Each sublayer is provided with 64 feature channels, and multiple relationship information is integrated through the message transmission mechanism to obtain the graph feature vector; Constructing a multi-branch feature extraction network for the energy consumption pattern clustering data, wherein the multi-branch feature extraction network includes a compressor unit feature branch, a condenser feature branch, and an evaporator feature branch, each branch includes two fully connected layers, and the number of neurons is 128 and 64, respectively, to obtain a device feature vector; Input the device feature vector into a time series feature extraction layer, wherein the time series feature extraction layer adopts a gated recurrent unit structure, includes an input gate, a reset gate, and an update gate, the hidden state dimension is set to 128, and the time series window length is 24, to obtain a time series feature vector; The graph feature vector and the time series feature vector are input into a cross attention module, the cross attention module is set with 8 attention heads, each with a dimension of 32, and adaptive fusion is performed by calculating the correlation weights between features to obtain multimodal fusion features; Inputting the multimodal fusion features into a feature enhancement network, the feature enhancement network comprises three residual blocks, each residual block is composed of two convolutional layers and a short-circuit connection, and a batch normalization layer and a ReLU activation function are set between the residual blocks to obtain an enhanced feature vector; The enhanced feature vector is input into an abnormal pattern recognition layer, wherein the abnormal pattern recognition layer includes an abnormal energy consumption sub-network, an abnormal temperature sub-network and an abnormal pressure sub-network, each sub-network is composed of three fully connected layers, and the number of neurons is 256, 128 and 64 respectively, to obtain an abnormal pattern vector; A comprehensive analysis is performed on the abnormal pattern vector, and the weight coefficients of energy consumption abnormality, temperature abnormality, and pressure abnormality are set to 0.4, 0.3, and 0.3, respectively. The abnormal energy consumption characteristic data of the equipment is obtained through weighted combination.

6. The energy management and recovery method for cold chain supply chain according to claim 5, characterized in that: The generating of the initial energy recovery scheme according to the abnormal energy consumption characteristic data of the equipment includes: A recycling priority matrix is ​​constructed for the abnormal energy consumption characteristic data of the equipment, and the abnormal degree is graded based on the energy consumption abnormality weight of 0.4, the temperature abnormality weight of 0.3 and the pressure abnormality weight of 0.3 to obtain abnormal priority data; An energy recovery type mapping table is established according to the abnormal priority data, and a corresponding relationship is established between the abnormal characteristics and three recovery modes: compressor waste heat recovery, condensation heat recovery, and system energy balance, to obtain recovery type data; Expanding the features of the recycling type data, constructing a feature enhancement matrix by introducing historical recycling efficiency indicators and equipment operation status indicators, and obtaining recycling feature data; The recovered feature data is input into the input layer of the Softmax regression classifier, and processed by two hidden layers, each layer is provided with 128 neurons and configured with BatchNorm and ReLU activation functions to obtain classification feature data; Generate a recovery strategy probability distribution according to the classification feature data, perform probability scoring on the compressor waste heat recovery parameters, condensation heat recovery parameters and system energy balance parameters to obtain strategy probability data, and construct a parameter configuration matrix based on the strategy probability data, including the value ranges of the compressor speed, condensation temperature and evaporation temperature, to obtain parameter configuration data; An operation constraint analysis is performed on the parameter configuration data, and a parameter adjustment step is set according to the equipment operation boundary conditions and system stability requirements to obtain operation parameter data. The operation parameter data is organized in execution order according to the system response timing and energy transfer link, and a control instruction sequence is constructed to obtain an initial energy recovery plan.

7. The energy management and recovery method for cold chain supply chain according to claim 6, characterized in that: The initial energy recovery scheme is applied to the cold chain system, operation efficiency data and load matching data are collected, and online iterative optimization is performed to obtain a target energy management scheme, including: The operating parameter data in the initial energy recovery scheme is sent to the control systems of the compressor unit, the condenser and the evaporator, and the compressor waste heat recovery data, the condensation heat recovery data and the system energy balance data are collected to obtain the first operating data; The first operation data is collected in sections according to three time scales of 5 minutes, 15 minutes and 30 minutes, and the dynamic change process of the equipment operation efficiency parameter and the load matching parameter is recorded to obtain the second operation data; Calculate the short-term operation efficiency index, the medium-term operation efficiency index and the long-term operation efficiency index respectively according to the second operation data, including the compressor energy efficiency ratio, the condenser heat exchange efficiency and the evaporator refrigeration coefficient, to obtain efficiency evaluation data; Performing a multi-scale load matching analysis on the efficiency evaluation data, calculating the deviations of the real-time load, hourly load and daily load from the set value, and performing a load dynamic balance calculation to obtain load matching data; Inputting the efficiency evaluation data and the load matching data into an adaptive optimization iterator, dynamically adjusting the energy recovery parameters by a stochastic gradient descent method with momentum term, and obtaining first optimization parameter data; Performing a hierarchical operation constraint check on the first optimization parameter data, and performing parameter correction according to the equipment safe operation boundary, system stability requirements, and energy balance constraints to obtain second optimization parameter data; Constructing a hierarchical control instruction sequence according to the second optimization parameter data, reconstructing the execution strategy according to the timing characteristics of the fast response layer, the mid-term adjustment layer and the slow optimization layer, and obtaining hierarchical control data; The hierarchical control data and the initial energy recovery plan are subjected to multi-objective fusion optimization, a closed-loop control process with a prediction-feedback-compensation mechanism is constructed, and a target energy management plan is obtained.

8. An energy management and recovery system for a cold chain supply chain, characterized in that: The method for energy management and recovery of a cold chain supply chain according to any one of claims 1 to 7 comprises: The acquisition module is used to collect and preprocess the operating parameters and environmental parameters of the cold chain system to obtain multi-dimensional feature data; A calculation module, used for performing feature extraction and local density distribution calculation on the multi-dimensional feature data to obtain energy consumption pattern clustering data; A construction module, used to construct a device operation knowledge graph based on the energy consumption pattern clustering data; An analysis module, used for performing abnormal feature analysis on the equipment operation knowledge graph and the energy consumption pattern clustering data through a hybrid neural network model to obtain abnormal feature data of equipment energy consumption; A generating module, used for generating an initial energy recovery plan according to the abnormal energy consumption characteristic data of the equipment; The optimization module is used to apply the initial energy recovery plan to the cold chain system, collect operation efficiency data and load matching data, and perform online iterative optimization to obtain a target energy management plan.

9. An energy management and recovery device for a cold chain supply chain, characterized in that: The energy management and recovery device of the cold chain supply chain includes: a memory and at least one processor, wherein the memory stores instructions; The at least one processor calls the instructions in the memory to enable the energy management and recovery device of the cold chain supply chain to execute the energy management and recovery method of the cold chain supply chain as described in any one of claims 1-7.

10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the energy management and recovery method for the cold chain supply chain as described in any one of claims 1-7 is implemented.

Images