Intelligent comprehensive energy management and control method and system based on big data platform
By constructing a smart integrated energy management and control system based on a big data platform, real-time monitoring and dynamic allocation of energy equipment have been achieved, solving the problems of low efficiency and untimely fault handling in traditional energy management methods, and improving the stability and security of the energy system.
Patent Information
- Application Number
- CN202410996832.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-24
- Publication Date
- 2024-11-22
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Traditional energy management methods lack systematicness and intelligence, resulting in low energy utilization efficiency. Monitoring the operating status of energy equipment relies on manual labor, making it difficult to detect and handle faults in a timely manner, leading to unstable energy supply and safety hazards.
Construct a smart integrated energy management and control system based on a big data platform. Through data acquisition, hierarchical data communication network and energy management and control model set, realize real-time monitoring and dynamic energy allocation of energy equipment. Use load forecasting, energy demand management and equipment failure prediction models for in-depth analysis and intelligent decision-making.
It improves energy efficiency, reduces energy costs, decreases equipment failure rate, ensures the continuity of energy supply and system stability, and enhances the accuracy of intelligent integrated energy management.
Smart Images

Figure CN119005884A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of big data energy management and control technology, and in particular to a smart integrated energy management and control method and system based on a big data platform. Background Art
[0002] With the continuous growth of energy demand and the complexity of energy structure, traditional energy management methods have been unable to meet the needs of modern energy systems. Traditional energy management methods usually rely on human experience and simple control strategies, lack of systematicity and intelligence, resulting in low energy utilization efficiency and serious energy waste. In addition, the operation status monitoring and maintenance of energy equipment rely on manual inspections, which makes it difficult to detect and handle equipment failures in a timely manner, resulting in unstable energy supply and potential safety hazards.
[0003] In modern energy systems, various energy equipment and facilities are widely distributed, involving multiple links such as power generation, transmission, distribution, and power consumption, and the amount of data is huge and complex. The existing energy management system lacks effective data collection and processing capabilities, and cannot fully and real-time grasp the operating status of energy equipment and energy usage. Data islands are common, and it is difficult to achieve data interoperability and sharing between systems, resulting in a low level of informationization and intelligence in energy management, affecting the overall efficiency and reliability of the energy system. Summary of the invention
[0004] The present application provides a smart integrated energy management and control method and system based on a big data platform, which is used to improve the accuracy of smart integrated energy management and control by adopting big data platform technology.
[0005] In a first aspect, the present application provides a smart integrated energy management and control method based on a big data platform, and the smart integrated energy management and control method based on a big data platform includes:
[0006] Build a smart integrated energy management and control platform and collect data from multiple energy devices to obtain target energy usage data and target device status data for each energy device;
[0007] A hierarchical data communication network is used to allocate a data transmission path for each energy device, and the target energy usage data and the target device status data are transmitted to a central computing node of the smart integrated energy management and control platform through the data transmission path;
[0008] Calling the energy management and control model set in the central computing node to perform energy management and control decision analysis on the target energy usage data and the target device status data of each energy device, respectively, to obtain an energy management and control decision set for each energy device;
[0009] Dynamically allocate energy to the multiple energy devices based on the energy management and control decision set, and generate corresponding target energy allocation strategies.
[0010] In a second aspect, the present application provides a smart integrated energy management and control system based on a big data platform, and the smart integrated energy management and control system based on a big data platform includes:
[0011] The collection module is used to build a smart integrated energy management and control platform and collect data from multiple energy devices to obtain the target energy usage data and target device status data of each energy device;
[0012] A transmission module, used to allocate a data transmission path for each energy device using a hierarchical data communication network, and transmit the target energy usage data and the target device status data to a central computing node of the smart integrated energy management and control platform through the data transmission path;
[0013] An analysis module is used to call the energy management and control model set in the central computing node, perform energy management and control decision analysis on the target energy usage data and target device status data of each energy device, and obtain an energy management and control decision set for each energy device;
[0014] A generation module is used to dynamically allocate energy to the multiple energy devices based on the energy management and control decision set and generate a corresponding target energy allocation strategy.
[0015] In the technical solution provided by this application, by constructing a smart integrated energy management and control platform, the operation data of multiple energy devices can be collected in real time, including energy usage data and equipment status data. The data acquisition module has a data preprocessing function, which can filter, clean and format the data in real time to ensure the accuracy and consistency of the data. It adopts a hierarchical data communication network design, which can support high-speed and reliable data transmission, and has the ability to resist network attacks to ensure the security of data transmission. The hierarchical design meets the data transmission requirements of different levels, supports remote data transmission, calls the energy management and control model set in the central computing node, and conducts in-depth analysis and intelligent decision-making on the collected energy data. Through multi-task deep learning models such as load prediction models, energy demand management models, and equipment fault prediction and diagnosis models, dynamic energy allocation and optimization control of energy equipment are realized. Based on the results of big data analysis, the system can monitor and warn energy equipment in real time, generate intelligent warning decision information, improve energy utilization efficiency, reduce energy consumption costs, reduce equipment failure rate, and improve the stability and safety of the energy system. When a device failure or abnormality is detected, the fault response mechanism can be quickly started to ensure the continuity of energy supply and the stable operation of the system, thereby improving the accuracy of smart integrated energy management and control. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] 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.
[0017] Figure 1 This is a schematic diagram of an embodiment of a smart integrated energy management and control method based on a big data platform in an embodiment of the present application;
[0018] Figure 2 This is a schematic diagram of an embodiment of an intelligent integrated energy management and control system based on a big data platform in an embodiment of the present application. DETAILED DESCRIPTION
[0019] The embodiment of the present application provides a method and system for intelligent integrated energy management and control based on a big data platform. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described here can be implemented in an order other than that illustrated or described here. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0020] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 In the embodiment of the present application, an embodiment of a smart integrated energy management and control method based on a big data platform includes:
[0021] Step S101: construct a smart integrated energy management and control platform and collect data from multiple energy devices to obtain target energy usage data and target device status data for each energy device;
[0022] It is understandable that the execution subject of the present application can be a smart integrated energy management and control system based on a big data platform, or a terminal or a server, which is not limited here. The present application embodiment is described by taking a server as the execution subject as an example.
[0023] Specifically, a central computing node is created based on a distributed cloud service, and a first edge node for each user terminal and a second edge node for each energy device are created. According to the distribution and functions of the central computing node, the first edge node and the second edge node, a smart integrated energy management and control platform is constructed, which can monitor and manage the operating status and energy usage of multiple energy devices in real time. Data is collected from multiple energy devices through the smart integrated energy management and control platform to obtain the initial energy usage data and initial device status data of each energy device. In the process of data collection, the smart integrated energy management and control platform uses its distributed architecture and data processing capabilities to ensure that the collected data can timely and accurately reflect the actual operation of each energy device. Based on the second edge node, the initial energy usage data and initial device status data of each energy device are cleaned and denoised, various noises and abnormal information are removed, and the standard energy usage data and standard device status data of each energy device are obtained. The abnormal data detection algorithm is used to detect and remove abnormal values for the standard energy usage data and standard device status data of each energy device. The abnormal data detection algorithm can identify abnormal values in the data and remove these abnormal values to ensure the accuracy and reliability of the data. After outlier detection and outlier removal, the target energy usage data and target device status data of each energy device are finally obtained.
[0024] The abnormal data detection algorithm is used to calculate the abnormal score of the standard energy usage data and the standard device status data of each energy device, evaluate the abnormal degree of each data point, obtain the first abnormal score value of each energy usage data point and the second abnormal score value of each device status data point, and quantitatively measure the abnormal degree of each data point. The first abnormal score threshold of the energy usage data is obtained, and the energy usage data points are judged to be abnormal by the threshold. By traversing and comparing the first abnormal score threshold and the first abnormal score value of each energy usage data point, multiple abnormal energy usage data points are identified. These abnormal data points are often significantly different from normal data points, which may be caused by sensor failure, data transmission error or other abnormal conditions. Similarly, for the device status data, a second abnormal score threshold is obtained, which is used to judge which device status data points are abnormal. By traversing and comparing the second abnormal score threshold and the second abnormal score value of each device status data point, multiple abnormal device status data points are identified. These data points may reflect abnormal conditions in the operation of the device, such as device failure or abnormal state. For multiple abnormal energy usage data points in the standard energy usage data, they are removed from the data set to obtain the target energy usage data of each energy device. By removing these abnormal data points, the accuracy and representativeness of the remaining data are ensured, and the interference of abnormal data on subsequent analysis and decision-making is avoided. For multiple abnormal device status data points in the standard device status data, they are removed from the data set to obtain the target device status data for each energy device.
[0025] Step S102: using a hierarchical data communication network to allocate a data transmission path for each energy device, and transmitting the target energy usage data and the target device status data to a central computing node of the smart integrated energy management and control platform through the data transmission path;
[0026] Specifically, a hierarchical data communication network structure is designed, which can effectively manage and transmit a large amount of data from multiple energy devices. In a hierarchical data communication network, it usually contains multiple layers, each of which corresponds to a specific data transmission and processing function. The bottom layer is the energy device layer, which is directly connected to each energy device and is responsible for preliminary data collection and transmission. Each energy device collects real-time energy usage data and device status data through built-in sensors and data collection modules, and transmits the data to the next layer through a local communication interface. The edge nodes of the edge computing layer are deployed close to the energy devices, receive data from the energy devices, and perform preliminary processing and filtering. The raw data is preprocessed by the edge nodes, such as data cleaning, data denoising, and preliminary data analysis. These preprocessing steps can significantly reduce data redundancy and noise, and improve data quality and transmission efficiency. At the same time, the edge nodes can also cache data to reduce data loss caused by network delays or transmission interruptions, thereby ensuring the reliability of data transmission. The data is transmitted to the distributed network layer. The distributed network layer is usually composed of multiple intermediate nodes, which are interconnected through a high-speed network to form a widely covered and efficient data transmission channel. At this layer, data can select the most effective transmission path through the optimal path selection algorithm to ensure that the data can be transmitted to the target node quickly and stably. The selection of the transmission path not only considers the transmission speed and bandwidth of the network, but also the load balancing and reliability of the network to prevent congestion and bottlenecks during data transmission. Finally, the data is transmitted to the highest layer, that is, the central computing node. The central computing node is usually deployed in a data center or cloud and has powerful computing and storage capabilities. The central computing node is responsible for receiving and integrating data from multiple edge nodes and performing in-depth analysis and processing. In the central computing node, the data is processed through advanced data analysis and machine learning algorithms to generate energy management decisions and optimization solutions. At the same time, the central computing node returns the processing results to the edge nodes and energy devices to ensure the closed-loop operation of the entire energy management system.
[0027] Step S103: calling the energy management and control model set in the central computing node, respectively performing energy management and control decision analysis on the target energy usage data and target device status data of each energy device, and obtaining an energy management and control decision set for each energy device;
[0028] Specifically, the energy management and control model set in the central computing node is called. These models integrate multiple analysis and prediction algorithms and can perform in-depth analysis on a large amount of complex data. The target energy usage data and target device status data of each energy device are respectively constructed to obtain the energy usage curve and device status curve of each energy device. These curves can intuitively reflect the usage and status changes of energy devices in different time periods. The energy usage curve and the device status curve are identified by curve feature points, and the key points in the curve are extracted. These key points usually include peak values, valley values, change inflection points, etc., which can reflect important events or status changes in the operation of energy devices. After identifying these feature points, the energy curve feature point set and the device curve feature point set of each energy device are obtained. The energy curve feature point set and the device curve feature point set are respectively normalized and feature encoded, and data of different scales are converted to a unified scale for subsequent analysis and processing. Feature encoding converts the normalized feature points into encoding vectors that can be input into the model, and obtains the energy usage encoding vector and the device status encoding vector of each energy device. The energy usage coding vector and the equipment status coding vector are vector-fused, and the two different coding vectors are merged into a comprehensive target fusion coding vector to fully reflect the operation status of the energy equipment. The target fusion coding vector of each energy device is input into the preset energy management and control model set. The energy management and control model set includes multiple models such as load forecasting model, energy demand management model, and equipment fault prediction and diagnosis model. The load forecasting model is used to predict future energy demand, the energy demand management model is used to optimize energy distribution, and the equipment fault prediction and diagnosis model is used to identify and predict equipment failures. By inputting the target fusion coding vector into these models, the advantages of the models are used to comprehensively analyze and predict the operation status of the energy equipment. Through the energy management and control model set, the target fusion coding vector of each energy device is analyzed for energy management and control decisions, and the energy management and control decision set of each energy device is obtained. The decision set contains specific suggestions and optimization plans for the operation of energy equipment, such as adjusting energy usage strategies and preventing equipment failures.
[0029] The energy curve feature point set of each energy device is subjected to feature normalization processing to normalize data of different scales into a unified range, usually to normalize data between 0 and 1. Through feature normalization, the dimension differences between different features are eliminated to ensure the effectiveness of subsequent processing. The normalized energy curve feature point set is converted into one-hot vector encoding. One-hot vector encoding is a way to convert categorical data into binary vectors so that each feature point has a unique representation in the encoding vector. Through one-hot vector encoding, the normalized feature points are converted into the initial use encoding vector. The device curve feature point set of each energy device is subjected to feature normalization processing to ensure that the device status data is within the same scale range. One-hot vector encoding conversion is performed to obtain the initial state encoding vector. The device status feature is encoded by one-hot vector to obtain a unique and effective representation. The initial encoding vector is subjected to principal component feature dimensionality reduction processing. Principal component analysis is a statistical technique that projects the original data into a new coordinate system through linear transformation to maximize the variance of the projected new coordinate system. Principal component feature dimensionality reduction processing of the initial use encoding vector can reduce the dimension of the data while retaining most of the information. Through dimensionality reduction processing, the energy usage encoding vector of each energy device is obtained. The initial state encoding vector is processed by principal component feature dimensionality reduction to reduce the dimension of the device state feature and retain key information. The device state encoding vector after dimensionality reduction can effectively represent the operating status of the device, which is convenient for subsequent analysis and processing.
[0030] The target fusion coding vector of each energy device is input into the load prediction model, which includes a spatiotemporal graph convolutional network and a first output layer. The spatiotemporal graph convolutional network is used to extract spatiotemporal features of the target fusion coding vector, and generate a load spatiotemporal feature vector by capturing the spatial and temporal features in the data. The load spatiotemporal feature vector makes a load decision through the first output layer to obtain the load decision information of each energy device. The target fusion coding vector of each energy device is input into the energy demand management model, which includes a long short-term memory network and a second output layer. The long short-term memory network can effectively process time series data, and generates an energy demand feature vector by extracting time series features from the target fusion coding vector. The energy demand feature vector makes an energy demand decision through the second output layer to obtain the energy demand decision information of each energy device. These decision information can help formulate energy supply strategies and ensure the balance of energy supply and demand. The target fusion coding vector of each energy device is input into the equipment fault prediction and diagnosis model, which includes a threshold recurrent network and a third output layer. The threshold recurrent network generates a state hidden feature vector by extracting hidden state features from the target fusion coding vector. The hidden feature vector is used to predict and diagnose equipment failures through the third output layer to obtain equipment failure decision information for each energy device. Equipment failure decision information can help to identify and predict equipment failures in a timely manner to ensure the normal operation and maintenance of the equipment. Load decision information, energy demand decision information, and equipment failure decision information are integrated to obtain the energy management and control decision set for each energy device.
[0031] Step S104: Dynamically allocate energy to multiple energy devices based on the energy management and control decision set, and generate corresponding target energy allocation strategies.
[0032] Specifically, the load decision information is used to predict the energy load demand of each device in the future. These prediction data can help understand the energy usage trend and peak load of each device in different time periods. Combined with the energy demand decision information, the energy allocation strategy is refined to ensure that the energy supply can meet the actual needs of each device. At the same time, the equipment failure decision information can provide the health status and failure prediction data of the equipment. Through this information, the possible failure risk of the equipment can be predicted in advance, so as to reserve sufficient redundancy in the energy allocation strategy to prevent insufficient energy supply caused by equipment failure. In order to achieve dynamic energy allocation, a real-time monitoring and adjustment mechanism is introduced to monitor the operating status and energy usage of each energy device in real time, and dynamically adjust according to the actual situation. For example, when the actual load of a certain device exceeds the predicted value, the system can immediately adjust the energy allocation strategy to increase the energy supply to the device to ensure its normal operation. Similarly, when the energy demand of a certain device is lower than expected, the system can reduce the energy supply to the device and allocate the excess energy to other devices with higher demand, so as to achieve efficient use of energy. Design an optimization algorithm to generate the optimal energy allocation strategy. The optimization algorithm can be based on linear programming, dynamic programming or other optimization techniques, and determine the optimal energy allocation plan by comprehensively calculating various data such as the energy demand, operating status and fault prediction of each device. The plan needs to meet the energy demand of each device, maximize energy utilization efficiency and reduce energy waste. At the same time, the optimization algorithm also needs to take into account factors such as equipment priority, failure risk and maintenance requirements to ensure that the energy allocation plan can take into account the stability and reliability of the system under any circumstances. The target energy allocation strategy generated by the optimization algorithm can realize dynamic energy allocation for multiple energy devices. According to the generated target energy allocation strategy, the energy supply of each device is adjusted as needed to ensure that each device operates in the best state. At the same time, based on real-time monitoring data, the energy allocation strategy is continuously optimized to ensure efficient and intelligent energy management.
[0033] Generate multiple corresponding target control instructions based on the target energy allocation strategy. These instructions include specific operations for each energy device, such as adjusting power output, changing operating mode, starting or stopping equipment, etc. These instructions are sent to each energy device through the smart integrated energy management and control platform. With the help of a hierarchical data communication network, ensure that the instructions can be transmitted to each device quickly and accurately. Through the hierarchical network, the control instructions are transmitted from the central computing node to each edge node, and then distributed to each energy device by the edge node. After receiving the target control instruction, each energy device responds to the instruction. Each device adjusts its operating parameters and status according to the specific requirements of the instruction to implement the target energy allocation strategy. During the execution process, the energy device monitors its operating status and execution effect in real time, and returns this information as the target execution result to the smart integrated energy management and control platform. These execution results can reflect the actual response of the device, help the platform understand the execution effect of the instruction, and provide data support for subsequent adjustments and optimizations. The smart integrated energy management and control platform evaluates and diagnoses the overall system. If some devices are found to have abnormalities or failures during the execution process, the platform generates corresponding fault response mechanisms. These mechanisms include fault identification, location, analysis, and solution formulation to ensure that the system can respond to various faults quickly and effectively and maintain stable operation. The fault response mechanism is sent to multiple user terminals based on the first edge node. Through the edge node, the fault response mechanism can be quickly transmitted to each user terminal, ensuring that users can understand the system's fault conditions and response plans in a timely manner.
[0034] In the embodiment of the present application, by constructing a smart integrated energy management and control platform, the operation data of multiple energy devices can be collected in real time, including energy usage data and equipment status data. The data acquisition module has a data preprocessing function, which can filter, clean and format the data in real time to ensure the accuracy and consistency of the data. It adopts a hierarchical data communication network design, which can support high-speed and reliable data transmission, and has the ability to resist network attacks to ensure the security of data transmission. The hierarchical design meets the data transmission requirements of different levels, supports remote data transmission, calls the energy management and control model set in the central computing node, and conducts in-depth analysis and intelligent decision-making on the collected energy data. Through multi-task deep learning models such as load prediction models, energy demand management models, and equipment fault prediction and diagnosis models, dynamic energy allocation and optimization control of energy equipment are realized. Based on the results of big data analysis, the system can monitor and warn energy equipment in real time, generate intelligent warning decision information, improve energy utilization efficiency, reduce energy consumption costs, reduce equipment failure rate, and improve the stability and safety of the energy system. When a device failure or abnormality is detected, the fault response mechanism can be quickly started to ensure the continuity of energy supply and the stable operation of the system, thereby improving the accuracy of smart integrated energy management and control.
[0035] In a specific embodiment, the process of executing step S101 may specifically include the following steps:
[0036] (1) creating a central computing node based on a distributed cloud service, and creating a first edge node for each user terminal and a second edge node for each energy device;
[0037] (2) Building a smart integrated energy management and control platform based on the central computing node, the first edge node, and the second edge node;
[0038] (3) Collect data from multiple energy devices through the smart integrated energy management and control platform to obtain the initial energy usage data and initial device status data of each energy device;
[0039] (4) performing data cleaning and data denoising on the initial energy usage data and initial device status data of each energy device based on the second edge node to obtain standard energy usage data and standard device status data of each energy device;
[0040] (5) An abnormal data detection algorithm is used to perform abnormal value detection and abnormal value removal on the standard energy usage data and standard equipment status data of each energy device, so as to obtain the target energy usage data and target equipment status data of each energy device.
[0041] Specifically, a central computing node is created on the distributed cloud service platform. The node is the core processing unit of the entire system and is responsible for high-intensity data computing, analysis and storage. The central computing node uses the elasticity and computing power of cloud computing to process data requests from a large number of energy devices and user terminals. A first edge node is created on each user terminal. These nodes are close to the user device and are responsible for the preliminary processing and caching of local data. The introduction of edge computing can significantly reduce the delay of data transmission and improve the real-time performance of data processing. At the same time, a second edge node is created on each energy device. These nodes are directly connected to the energy device and are responsible for the real-time collection and preliminary processing of device data. The deployment of the second edge node can ensure the accuracy and timeliness of data collection while reducing the burden on the central computing node. A comprehensive smart energy management and control platform is constructed based on the central computing node, the first edge node and the second edge node. The platform uses a distributed architecture to organically combine the central computing node, the user terminal and the energy device to achieve efficient data transmission and processing. Data is collected from multiple energy devices through the smart integrated energy management and control platform. The second edge node collects the initial energy usage data and initial device status data of each energy device in real time, which includes the power consumption, operating status, environmental parameters, etc. of the device. After the initial data is collected, the data is cleaned and denoised. The purpose of data cleaning is to remove invalid information and erroneous data in the data to improve the accuracy and reliability of the data. Data denoising is to remove random noise and interference signals in the data to obtain a smoother and more reliable data signal. For example, using a moving average filter to remove data noise, for a certain moment of energy usage data x i , its smoothed value can be calculated by the following formula:
[0042]
[0043] Among them, x′ i is the smoothed data, N is the window size of the moving average filter, x j is the original data. In this way, the high-frequency noise in the data can be effectively removed to obtain smoother and more reliable standard energy usage data and standard equipment status data. The abnormal data detection algorithm is used to detect and remove abnormal values from the standard energy usage data and standard equipment status data of each energy device. The purpose of abnormal data detection is to identify abnormal points in the data, which may be caused by sensor failure, data transmission errors, etc.
[0044] In a specific embodiment, the execution step uses an abnormal data detection algorithm to perform abnormal value detection and abnormal value removal on the standard energy usage data and standard device status data of each energy device, and the process of obtaining the target energy usage data and target device status data of each energy device can specifically include the following steps:
[0045] (1) performing abnormality score calculations on the standard energy usage data and the standard device status data of each energy device respectively by using an abnormal data detection algorithm to obtain a first abnormality score value for each energy usage data point and a second abnormality score value for each device status data point;
[0046] (2) obtaining a first abnormality scoring threshold of the energy usage data, and performing a traversal comparison between the first abnormality scoring threshold and the first abnormality scoring value to obtain a plurality of abnormal energy usage data points;
[0047] (3) obtaining a second abnormality scoring threshold of the device status data, and performing traversal comparison on the second abnormality scoring threshold and the second abnormality scoring value to obtain a plurality of abnormal device status data points;
[0048] (4) Removing multiple abnormal energy usage data points from each standard energy usage data to obtain target energy usage data for each energy device, and removing multiple abnormal device status data points from each standard device status data to obtain target device status data for each energy device.
[0049] Specifically, an abnormal data detection algorithm is used to calculate abnormal scores for the standard energy usage data and standard device status data of each energy device, and the score values can be calculated by statistical methods or machine learning algorithms. For example, the z-score method is used to calculate the abnormal score value. The z-score method measures the degree of deviation of a data point by comparing each data point with the mean and standard deviation of the data set. For each energy usage data point, its first abnormal score value is calculated, and for each device status data point, its second abnormal score value is calculated, thereby quantifying the degree of abnormality of each data point. Obtain a first abnormal score threshold for energy usage data and a second abnormal score threshold for device status data. These thresholds can be determined based on experience or by statistical methods. Usually, the threshold of the z-score method is set to 3, that is, if |z i |>3, then the data point is considered an outlier. Then the energy usage data point x i is considered abnormal; if Then the device status data point x iis considered abnormal. After obtaining the threshold, the abnormal score value and threshold of each data point are traversed and compared to identify abnormal data points. For example, for energy usage data, the first abnormal score value of each data point is traversed and compared with the first abnormal score threshold, and all abnormal energy usage data points are marked. Similarly, for device status data, the second abnormal score value of each data point is traversed and compared with the second abnormal score threshold, and all abnormal device status data points are marked. After all abnormal data points are identified, these abnormal data points are removed from the standard energy usage data and the standard device status data respectively to obtain the target energy usage data and the target device status data.
[0050] In a specific embodiment, the process of executing step S103 may specifically include the following steps:
[0051] (1) Call the energy management model set in the central computing node;
[0052] (2) constructing curves for the target energy usage data and the target device status data of each energy device respectively, and obtaining an energy usage curve and a device status curve for each energy device;
[0053] (3) Respectively identify curve feature points of the energy usage curve and the equipment status curve to obtain an energy curve feature point set and an equipment curve feature point set for each energy device;
[0054] (4) performing feature normalization and feature encoding on the energy curve feature point set and the device curve feature point set respectively, and obtaining the energy usage encoding vector and the device state encoding vector of each energy device;
[0055] (5) performing vector fusion on the energy usage coding vector and the device status coding vector to obtain a target fusion coding vector for each energy device;
[0056] (6) inputting the target fusion coding vector of each energy device into a preset energy management and control model set, the energy management and control model set including: a load prediction model, an energy demand management model, and an equipment fault prediction and diagnosis model;
[0057] (7) The target fusion coding vector of each energy device is separately fused through the energy control model set to perform energy control decision analysis, and obtain the energy control decision set for each energy device.
[0058] Specifically, the energy management and control model set in the central computing node is called. The energy management and control model set includes multiple models such as load prediction model, energy demand management model and equipment fault prediction and diagnosis model. Curves are constructed for the target energy usage data and target equipment status data of each energy device respectively. The energy usage curve and equipment status curve of each energy device are obtained by interpolating and smoothing the time series data. The energy usage curve reflects the energy consumption of the equipment in different time periods, while the equipment status curve shows the operating status and health status of the equipment. After obtaining the energy usage curve and equipment status curve, feature points of these curves are identified. Feature points refer to key nodes in the curve, such as peaks, valleys, turning points, etc. Through feature point identification, important information in the curve is extracted to form an energy curve feature point set and an equipment curve feature point set. Feature point identification can use a derivative-based method, that is, by calculating the first-order derivative and second-order derivative of the curve, the extreme points and change points of the curve are found. The energy curve feature point set and the device curve feature point set are feature normalized and feature encoded to convert data of different scales to the same scale. 0-1 normalization or z-score normalization is usually used. The normalized data is convenient for subsequent analysis and processing. Feature encoding is to convert feature points into numerical vectors for input into the model. For example, the feature points are converted into encoding vectors by using methods such as one-hot encoding or binary encoding. Through feature normalization and feature encoding, the energy usage encoding vector and the device status encoding vector of each energy device are obtained. The two encoding vectors are vector fused. The encoding vectors from two different sources are merged into a comprehensive vector. The target fused encoding vector obtained can comprehensively reflect the usage and operation status of the energy device. The target fused encoding vector of each energy device is input into the preset energy management and control model set. The energy management and control model set includes load forecasting model, energy demand management model and equipment fault prediction and diagnosis model. The load forecasting model is used to predict future energy demand. The spatiotemporal graph convolutional network is used to extract spatiotemporal features, and the load forecasting result is obtained through the output layer. The energy demand management model is mainly used to optimize energy distribution, using long short-term memory networks to extract time series features and obtain energy demand decisions through the output layer. The equipment failure prediction and diagnosis model is mainly used to predict and diagnose equipment failures, using threshold recurrent networks to extract hidden state features and obtain equipment failure prediction results through the output layer.
[0059] In a specific embodiment, the execution step performs feature normalization and feature encoding on the energy curve feature point set and the device curve feature point set of each energy device to obtain the energy usage encoding vector and the device state encoding vector of each energy device, which may specifically include the following steps:
[0060] (1) Perform feature normalization and one-hot vector encoding conversion on the energy curve feature point set of each energy device to obtain the initial usage encoding vector;
[0061] (2) Perform feature normalization and one-hot vector encoding conversion on the energy curve feature point set of each energy device to obtain the initial state encoding vector;
[0062] (3) Performing principal component feature dimensionality reduction processing on the initial usage coding vector to obtain the energy usage coding vector of each energy device;
[0063] (4) Perform principal component feature dimensionality reduction processing on the initial state coding vector to obtain the device state coding vector of each energy device.
[0064] Specifically, the energy curve feature point set of each energy device is normalized, and data of different scales are converted to the same scale to eliminate the influence of dimension, so that the influence of different features is balanced. Normalization can use the 0-1 normalization method to scale the data to the range of 0 to 1. The normalized data is converted into one-hot vector encoding. One-hot vector encoding is a way to convert categorical data into binary vectors. In one-hot encoding, each feature point is represented as a unique binary vector, in which only one position is 1 and the rest are 0. For example, assuming that the feature points include three categories A, B, and C, A can be encoded as [1, 0, 0], B is encoded as [0, 1, 0], and C is encoded as [0, 0, 1]. Through this encoding method, the normalized feature points are represented as one-hot encoded vectors. The feature point set of the equipment status is normalized. For example, assuming that the status data feature points of a certain equipment include running time, downtime, and number of failures, the data after normalization are 0.8, 0.1, and 0.05. Perform one-hot vector encoding conversion, assuming that the device status feature points include three states: normal, warning, and fault, which are encoded as [1, 0, 0], [0, 1, 0], and [0, 0, 1] respectively. Perform principal component feature dimensionality reduction processing on the initial use encoding vector and the initial state encoding vector. Principal component analysis is a commonly used dimensionality reduction technique that projects the original data into a new coordinate system through linear transformation to maximize the variance of the projected new coordinate system. Principal component feature dimensionality reduction processing includes the following aspects: Center the data, that is, adjust the mean of the data to 0. The formula is as follows:
[0065] x′=x-μ;
[0066] Where x′ is the centered data, x is the original data, and μ is the mean of the data. Calculate the covariance matrix of the centered data. The covariance matrix reflects the linear relationship between the various dimensions of the data. The formula is as follows:
[0067]
[0068] Where Σ is the covariance matrix, n is the number of samples of the data, and x′ i is the centralized data of the i-th sample. Perform eigenvalue decomposition on the covariance matrix to obtain eigenvalues and eigenvectors. The eigenvector represents the direction of the new coordinate axis, and the eigenvalue represents the variance size on each eigenvector. The formula is as follows:
[0069] Σ=VΛV T ;
[0070] Where V is the eigenvector matrix and Λ is the diagonal eigenvalue matrix. Select the eigenvectors corresponding to the first k largest eigenvalues to form the projection matrix W. Project the data to the new subspace to obtain the reduced-dimensional data matrix. The formula is as follows:
[0071] Z = X·W;
[0072] Among them, Z is the data matrix after dimensionality reduction, X is the original data matrix, and W is the projection matrix.
[0073] In a specific embodiment, the execution step performs energy management decision analysis on the target fusion coding vector of each energy device through the energy management model set to obtain the energy management decision set of each energy device, which may specifically include the following steps:
[0074] (1) Inputting the target fusion coding vector of each energy device into the load prediction model respectively, the load prediction model includes a spatiotemporal graph convolutional network and a first output layer; extracting spatiotemporal features of the target fusion coding vector through the spatiotemporal graph convolutional network to obtain a load spatiotemporal feature vector, and performing load decision on the load spatiotemporal feature vector through the first output layer to obtain load decision information of each energy device;
[0075] (2) respectively inputting the target fusion coding vector of each energy device into the energy demand management model, the energy demand management model including the long short-term memory network and the second output layer; extracting the time series features of the target fusion coding vector through the long short-term memory network to obtain the energy demand feature vector, and performing energy demand decision on the energy demand feature vector through the second output layer to obtain the energy demand decision information of each energy device;
[0076] (3) respectively inputting the target fusion coding vector of each energy device into the device fault prediction and diagnosis model, the device fault prediction and diagnosis model including a threshold recurrent network and a third output layer; extracting hidden state features of the target fusion coding vector through the threshold recurrent network to obtain a state hidden feature vector, and performing device fault prediction and diagnosis on the state hidden feature vector through the third output layer to obtain device fault decision information of each energy device;
[0077] (4) Integrate the load decision information, energy demand decision information, and equipment failure decision information to obtain the energy management and control decision set for each energy device.
[0078] Specifically, the target fusion coding vector of each energy device is input into the load prediction model. The load prediction model is processed by the spatiotemporal graph convolutional network. The spatiotemporal graph convolutional network can effectively capture the dependencies in the time and space dimensions, thereby extracting spatiotemporal features. The spatiotemporal graph convolutional network takes the target fusion coding vector as input, processes it through multiple spatiotemporal convolutional layers, and extracts the load spatiotemporal feature vector. The formula is as follows:
[0079]
[0080] Among them, H (l) represents the feature matrix of the lth layer, A k represents the k-th order adjacency matrix, D represents the degree matrix, W k Represents the convolution kernel weight matrix, and σ represents the activation function. Through these operations, the load spatiotemporal feature vector is obtained, and then the load decision is made through the first output layer to obtain the load decision information of each energy device. The target fusion coding vector of each energy device is input into the energy demand management model. The energy demand management model uses a long short-term memory network for processing to capture the long-term dependencies in the time series data and extract the timing features. The long short-term memory network takes the target fusion coding vector as input, processes it through multiple long short-term memory network layers, and extracts the energy demand feature vector. The formula is as follows:
[0081] h t =σ(W h [h t-1 ,x t ]+b h );
[0082] Among them, h t represents the hidden state at the current moment, x t represents the input at the current moment, W h represents the weight matrix, b h Represents the bias vector, and σ represents the activation function. Through these operations, the energy demand feature vector is obtained, and then the energy demand decision is made through the second output layer to obtain the energy demand decision information of each energy device. Subsequently, the target fusion coding vector of each energy device is input into the equipment fault prediction and diagnosis model. The equipment fault prediction and diagnosis model is processed by a threshold recurrent network. The threshold recurrent network can effectively capture the short-term dependencies and hidden state features in time series data, thereby extracting hidden state features. The threshold recurrent network takes the target fusion coding vector as input, processes it through multiple threshold recurrent network layers, and extracts the state hidden feature vector. The formula is as follows:
[0083] h t =(1-z t )⊙h t-1 +z t ⊙tanh(W h [r t ⊙h t-1 ,x t ]+b h );
[0084] Among them, h t represents the hidden state at the current moment, z t represents the update gate, r t represents the reset gate, x t represents the input at the current moment, W h represents the weight matrix, b h Represents the bias vector, ⊙ represents the bitwise multiplication of the elements, and tanh represents the hyperbolic tangent activation function. Through these operations, the state hidden feature vector is obtained, and then the equipment failure prediction and diagnosis are performed through the third output layer to obtain the equipment failure decision information of each energy device. The load decision information, energy demand decision information and equipment failure decision information are integrated to obtain the energy management and control decision set of each energy device. The information integration process can adopt the weighted average method to weight the decision results of different models to form the final energy management and control decision.
[0085] In a specific embodiment, the above-mentioned intelligent integrated energy management and control method based on the big data platform also includes the following steps:
[0086] (1) Generate multiple corresponding target control instructions based on the target energy allocation strategy, and issue the multiple target control instructions to multiple energy devices through the intelligent integrated energy management and control platform;
[0087] (2) Respond to and control multiple target control instructions through multiple energy devices, and return the corresponding target execution results to the smart integrated energy management and control platform;
[0088] (3) Generate a corresponding fault response mechanism through the intelligent integrated energy management and control platform, and send the fault response mechanism to multiple user terminals based on the first edge node.
[0089] Specifically, multiple target control instructions are generated based on the target energy allocation strategy. The target energy allocation strategy is generated by optimizing the energy demand of energy equipment and the overall energy supply of the system to ensure that energy equipment can achieve efficient use of energy while meeting the demand. The smart integrated energy management and control platform generates corresponding control instructions based on the energy demand decision information, load prediction information and equipment failure prediction information of each device. These control instructions include switching operations, power adjustment, and operation mode switching of the equipment. For example, if the target energy allocation strategy requires an industrial boiler to reduce power usage during peak hours, the generated control instructions may include specific operations such as reducing boiler fuel supply and adjusting steam output. The target control instructions are issued to each energy device through the smart integrated energy management and control platform. As the core dispatching unit of the system, the smart integrated energy management and control platform is responsible for transmitting control instructions to each device through the network. Using the Internet of Things technology, the control instructions are sent to the control unit of the device through wireless transmission or wired transmission. For example, the control instructions are transmitted to a transformer device in a smart grid system through the platform, requiring it to adjust the load during a specific period to cope with fluctuations in power demand. After each energy device receives the control instruction, the control unit inside the device will parse and execute the instruction. Energy equipment responds to control according to control instructions and adjusts its operating status. For example, industrial boilers reduce fuel supply and adjust steam output according to received control instructions, thereby achieving power regulation. At this time, the sensors and control modules of the equipment will monitor the execution results in real time and feed back these data as target execution results to the smart integrated energy management and control platform. Through this closed-loop control mechanism, the platform can understand the response of each device and the actual execution effect in real time. The corresponding fault response mechanism is generated through the smart integrated energy management and control platform. When the platform detects that some equipment has an abnormality or failure in the process of executing control instructions, it will generate a corresponding fault response mechanism. These mechanisms include fault identification, diagnosis, analysis and corresponding processing solutions. For example, when the platform detects that a transformer device has an overload fault when executing a load adjustment instruction, the platform will generate a fault response mechanism, including specific measures such as reducing load and switching to backup power supply. Subsequently, the smart integrated energy management and control platform will send the fault response mechanism to multiple user terminals through the first edge node. As a computing and storage unit close to the user end, the first edge node can quickly respond to the platform's instructions and pass the fault response mechanism to the user terminal in a timely manner. For example, the energy management center of the industrial park, as a user terminal, receives the fault response mechanism issued by the platform through the edge node, promptly understands the transformer fault situation and takes corresponding measures, such as dispatching maintenance personnel and starting backup equipment.
[0090] The above describes the intelligent integrated energy management method based on the big data platform in the embodiment of the present application. The following describes the intelligent integrated energy management system based on the big data platform in the embodiment of the present application. Figure 2 In the embodiment of the present application, an embodiment of a smart integrated energy management and control system based on a big data platform includes:
[0091] The collection module 201 is used to build a smart integrated energy management and control platform and collect data from multiple energy devices to obtain target energy usage data and target device status data for each energy device;
[0092] Transmission module 202, used to allocate a data transmission path for each energy device using a hierarchical data communication network, and transmit the target energy usage data and target device status data to a central computing node of the smart integrated energy management and control platform through the data transmission path;
[0093] The analysis module 203 is used to call the energy management and control model set in the central computing node, perform energy management and control decision analysis on the target energy usage data and target device status data of each energy device, and obtain an energy management and control decision set for each energy device;
[0094] The generation module 204 is used to dynamically allocate energy to multiple energy devices based on the energy management and control decision set and generate a corresponding target energy allocation strategy.
[0095] Through the collaboration of the above components, by building a smart integrated energy management and control platform, the operation data of multiple energy devices can be collected in real time, including energy usage data and equipment status data. The data acquisition module has the function of data preprocessing, which can filter, clean and format the data in real time to ensure the accuracy and consistency of the data. The hierarchical data communication network design can support high-speed and reliable data transmission, and has the ability to resist network attacks to ensure the security of data transmission. The hierarchical design meets the data transmission requirements of different levels, supports remote data transmission, calls the energy management and control model set in the central computing node, and conducts in-depth analysis and intelligent decision-making on the collected energy data. Through multi-task deep learning models such as load prediction model, energy demand management model and equipment fault prediction and diagnosis model, dynamic energy allocation and optimization control of energy equipment are realized. Based on the results of big data analysis, the system can monitor and warn energy equipment in real time, generate intelligent warning decision information, improve energy utilization efficiency, reduce energy consumption costs, reduce equipment failure rate, and improve the stability and safety of the energy system. When equipment failure or abnormality is detected, the fault response mechanism can be quickly activated to ensure the continuity of energy supply and the stable operation of the system, thereby improving the accuracy of smart integrated energy management and control.
[0096] 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.
[0097] 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, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform 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 and other media that can store program codes.
[0098] 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 smart integrated energy management and control method based on a big data platform, characterized in that: The method comprises: Build a smart integrated energy management and control platform and collect data from multiple energy devices to obtain target energy usage data and target device status data for each energy device; A hierarchical data communication network is used to allocate a data transmission path for each energy device, and the target energy usage data and the target device status data are transmitted to a central computing node of the smart integrated energy management and control platform through the data transmission path; Calling the energy management and control model set in the central computing node to perform energy management and control decision analysis on the target energy usage data and target device status data of each energy device, respectively, to obtain an energy management and control decision set for each energy device; Dynamically allocate energy to the multiple energy devices based on the energy management and control decision set, and generate corresponding target energy allocation strategies.
2. The intelligent integrated energy management and control method based on the big data platform according to claim 1 is characterized in that: The smart integrated energy management and control platform is constructed and data is collected from multiple energy devices to obtain target energy usage data and target device status data for each energy device, including: Creating a central computing node based on a distributed cloud service, and creating a first edge node for each user terminal and a second edge node for each energy device; Building a smart integrated energy management and control platform according to the central computing node, the first edge node and the second edge node; Data of multiple energy devices are collected through the smart integrated energy management and control platform to obtain initial energy usage data and initial device status data of each energy device; Based on the second edge node, the initial energy usage data and the initial device status data of each energy device are cleaned and denoised to obtain standard energy usage data and standard device status data of each energy device; The abnormal data detection algorithm is used to perform abnormal value detection and abnormal value removal on the standard energy usage data and standard equipment status data of each energy device, so as to obtain the target energy usage data and target equipment status data of each energy device.
3. The intelligent integrated energy management and control method based on the big data platform according to claim 2 is characterized in that: The abnormal data detection algorithm is used to perform abnormal value detection and abnormal value removal on the standard energy usage data and standard device status data of each energy device, respectively, to obtain the target energy usage data and target device status data of each energy device, including: Anomaly scoring is performed on the standard energy usage data and the standard device status data of each energy device respectively by using an abnormal data detection algorithm to obtain a first abnormality scoring value for each energy usage data point and a second abnormality scoring value for each device status data point; Obtaining a first abnormality scoring threshold of energy usage data, and traversing and comparing the first abnormality scoring threshold with the first abnormality scoring value to obtain a plurality of abnormal energy usage data points; Acquire a second abnormality scoring threshold of the device status data, and perform traversal comparison on the second abnormality scoring threshold and the second abnormality scoring value to obtain a plurality of abnormal device status data points; Multiple abnormal energy usage data points in each standard energy usage data are removed respectively to obtain target energy usage data for each energy device, and multiple abnormal device status data points in each standard device status data are removed respectively to obtain target device status data for each energy device.
4. The intelligent integrated energy management and control method based on the big data platform according to claim 1 is characterized in that: The energy management and control model set in the central computing node is called to perform energy management and control decision analysis on the target energy usage data and the target device status data of each energy device, respectively, to obtain an energy management and control decision set for each energy device, including: Calling the energy management and control model set in the central computing node; Constructing curves for target energy usage data and target device status data of each energy device respectively, and obtaining energy usage curves and device status curves of each energy device; Respectively identifying curve feature points of the energy usage curve and the equipment status curve to obtain an energy curve feature point set and an equipment curve feature point set for each energy equipment; Respectively performing feature normalization and feature encoding on the energy curve feature point set and the device curve feature point set to obtain an energy usage encoding vector and a device state encoding vector for each energy device; Performing vector fusion on the energy usage coding vector and the device status coding vector to obtain a target fusion coding vector for each energy device; Inputting the target fusion coding vector of each energy device into a preset energy management and control model set respectively, wherein the energy management and control model set includes: a load prediction model, an energy demand management model, and an equipment fault prediction and diagnosis model; The energy management and control model set is used to perform energy management and control decision analysis on the target fusion coding vector of each energy device to obtain an energy management and control decision set for each energy device.
5. The intelligent integrated energy management and control method based on the big data platform according to claim 4 is characterized in that: The step of normalizing and encoding the energy curve feature point set and the device curve feature point set of each energy device to obtain the energy usage encoding vector and the device state encoding vector of each energy device includes: Perform feature normalization and one-hot vector encoding conversion on the energy curve feature point set of each energy device to obtain the initial usage encoding vector; Perform feature normalization and one-hot vector encoding conversion on the energy curve feature point set of each energy device to obtain the initial state encoding vector; Performing principal component feature dimensionality reduction processing on the initial usage coding vector to obtain an energy usage coding vector for each energy device; The initial state coding vector is subjected to principal component feature dimensionality reduction processing to obtain a device state coding vector for each energy device.
6. The intelligent integrated energy management and control method based on the big data platform according to claim 4 is characterized in that: The energy management and control decision analysis is performed on the target fusion coding vector of each energy device through the energy management and control model set to obtain an energy management and control decision set for each energy device, including: The target fusion coding vector of each energy device is respectively input into the load prediction model, and the load prediction model includes a spatiotemporal graph convolutional network and a first output layer; the spatiotemporal feature of the target fusion coding vector is extracted through the spatiotemporal graph convolutional network to obtain a load spatiotemporal feature vector, and the load decision is made on the load spatiotemporal feature vector through the first output layer to obtain load decision information of each energy device; The target fusion coding vector of each energy device is respectively input into the energy demand management model, and the energy demand management model includes a long short-term memory network and a second output layer; the target fusion coding vector is subjected to time series feature extraction through the long short-term memory network to obtain an energy demand feature vector, and the energy demand feature vector is subjected to energy demand decision making through the second output layer to obtain energy demand decision information of each energy device; The target fusion coding vector of each energy device is respectively input into the device fault prediction and diagnosis model, and the device fault prediction and diagnosis model includes a threshold recurrent network and a third output layer; the hidden state feature of the target fusion coding vector is extracted through the threshold recurrent network to obtain a state hidden feature vector, and the device fault prediction and diagnosis is performed on the state hidden feature vector through the third output layer to obtain the device fault decision information of each energy device; The load decision information, the energy demand decision information and the equipment failure decision information are integrated to obtain an energy management and control decision set for each energy device.
7. The intelligent integrated energy management and control method based on the big data platform according to claim 2 is characterized in that: The method further comprises: Generate a corresponding plurality of target control instructions based on the target energy allocation strategy, and issue the plurality of target control instructions to the plurality of energy devices through the smart integrated energy management and control platform; Responding to and controlling the multiple target control instructions through the multiple energy devices, and returning the corresponding target execution results to the smart integrated energy management and control platform; A corresponding fault response mechanism is generated through the smart integrated energy management and control platform, and the fault response mechanism is sent to multiple user terminals based on the first edge node.
8. A smart integrated energy management and control system based on a big data platform, characterized in that: The system comprises: The collection module is used to build a smart integrated energy management and control platform and collect data from multiple energy devices to obtain the target energy usage data and target device status data of each energy device; A transmission module, used to allocate a data transmission path for each energy device using a hierarchical data communication network, and transmit the target energy usage data and the target device status data to a central computing node of the smart integrated energy management and control platform through the data transmission path; An analysis module is used to call the energy management and control model set in the central computing node, perform energy management and control decision analysis on the target energy usage data and target device status data of each energy device, and obtain an energy management and control decision set for each energy device; A generation module is used to dynamically allocate energy to the multiple energy devices based on the energy management and control decision set and generate a corresponding target energy allocation strategy.
9. The intelligent integrated energy management and control system based on the big data platform according to claim 8 is characterized in that: The acquisition module is specifically used for: Creating a central computing node based on a distributed cloud service, and creating a first edge node for each user terminal and a second edge node for each energy device; Building a smart integrated energy management and control platform according to the central computing node, the first edge node and the second edge node; Data of multiple energy devices are collected through the smart integrated energy management and control platform to obtain initial energy usage data and initial device status data of each energy device; Based on the second edge node, the initial energy usage data and the initial device status data of each energy device are cleaned and denoised to obtain standard energy usage data and standard device status data of each energy device; The abnormal data detection algorithm is used to perform abnormal value detection and abnormal value removal on the standard energy usage data and standard equipment status data of each energy device, so as to obtain the target energy usage data and target equipment status data of each energy device.
10. The intelligent integrated energy management and control system based on the big data platform according to claim 8 is characterized in that: The analysis module is specifically used for: Calling the energy management and control model set in the central computing node; Constructing curves for target energy usage data and target device status data of each energy device respectively, and obtaining energy usage curves and device status curves of each energy device; Respectively identifying curve feature points of the energy usage curve and the equipment status curve to obtain an energy curve feature point set and an equipment curve feature point set for each energy equipment; Respectively performing feature normalization and feature encoding on the energy curve feature point set and the device curve feature point set to obtain an energy usage encoding vector and a device state encoding vector for each energy device; Performing vector fusion on the energy usage coding vector and the device status coding vector to obtain a target fusion coding vector for each energy device; Inputting the target fusion coding vector of each energy device into a preset energy management and control model set respectively, wherein the energy management and control model set includes: a load prediction model, an energy demand management model, and an equipment fault prediction and diagnosis model; The energy management and control model set is used to perform energy management and control decision analysis on the target fusion coding vector of each energy device to obtain an energy management and control decision set for each energy device.
Citation Information
Patent Citations
Smart park energy comprehensive management and control platform based on data analysis
CN114298470A
Intelligent Internet of Things energy management platform system
CN117240887A
Centralized control platform management system based on optimization algorithm
CN117830029A
Energy management platform based on Internet of Things
CN117933660A
Cited By
Energy storage on-site monitoring method and system of energy storage power station
CN121077084A
