An energy collection monitoring system based on big data

Abstract

This application discloses a big data-based energy acquisition and monitoring system, belonging to the field of new energy equipment monitoring technology. It includes a data acquisition module that installs edge gateways on each new energy power generation device within the power grid to collect power generation data and status data of device components. A task allocation module establishes a regional gateway network among all edge gateways, constructs a dynamic task allocation model and deploys it to the edge gateways. Using the dynamic task allocation model, it selects multiple core gateways from all edge gateways and assigns different task plans to different core gateways. A task execution module enables the core gateways to perform point cloud map construction, fault diagnosis, lifespan prediction, and strategy generation according to the assigned task plans, thereby realizing the monitoring and management of new energy power generation equipment. This system achieves balanced task allocation among core gateways, improves the overall computing power utilization and task execution efficiency of the system, accurately identifies fault types, locates faulty components, and maintains the overall output stability of the power grid.

Description

Technical Field

[0001] This application belongs to the field of new energy equipment monitoring technology, specifically, it relates to an energy acquisition and monitoring system based on big data. Background Technology

[0002] With the rapid development of new energy power generation technologies, such as the widespread application of renewable energy sources like solar and wind power, energy acquisition and monitoring systems are playing an increasingly important role in power grid management. To ensure the stable operation of new energy power generation systems, it is necessary to collect and monitor the power generation data and component status of power generation equipment in real time. Through data analysis, fault diagnosis, life prediction, and operation strategy optimization can be achieved, which is of great significance to the safe and efficient operation of the power grid.

[0003] Current monitoring systems for new energy equipment typically deploy sensors on the power generation equipment to monitor its operating status, detect faults, and predict its lifespan. However, existing technologies only monitor and manage basic parameters such as current and voltage of the new energy equipment. When a fault occurs, they cannot accurately locate or predict the fault. There is a lack of collaborative analysis and energy efficiency optimization from a cluster perspective among the massive amounts of collected monitoring data, and a lack of deep fusion and feature extraction of multi-source data. This results in low fault detection accuracy, high false alarm rate, and an inability to accurately locate faulty components and predict remaining lifespan. Furthermore, when a fault occurs, fixed rules or simple feedback control are usually used, making it difficult to dynamically adjust the power generation equipment based on real-time environmental characteristics and equipment status, thus failing to effectively maintain the overall stability of the power grid. Summary of the Invention

[0004] To address the aforementioned problems and technical deficiencies, this application adopts the following technical solution: an energy acquisition and monitoring system based on big data, characterized in that it includes:

[0005] The data acquisition module is used to install an edge gateway on each new energy power generation device in the power grid to collect the power generation data and status data of the device components;

[0006] The task allocation module is used to build a regional gateway network among all edge gateways, construct a dynamic task allocation model and deploy it to the edge gateways, use the dynamic task allocation model to select multiple core gateways from all edge gateways, and allocate different task plans to different core gateways.

[0007] The task execution module enables the core gateway to perform point cloud map construction, fault diagnosis, lifespan prediction, and strategy generation according to the assigned task plan, thereby realizing the monitoring and management of new energy power generation equipment.

[0008] Preferably, the edge gateway has a built-in data compression algorithm to perform data cleaning, caching and compression. It adaptively adjusts the sampling frequency and compression ratio according to the rate of change of the collected data and the quality of the communication link. When communication with the regional gateway network is interrupted, it caches key data and prioritizes the transmission of abnormal data fragments after recovery.

[0009] Preferably, the task schedule allocation to the core gateway includes:

[0010] A dynamic task allocation model is used to select core gateways based on the amount of data collected, regional location, and number of gateways. Different weights are assigned to the amount of data collected, regional location concentration, and gateway computing load as evaluation indicators. A comprehensive score is calculated for each edge gateway, and the top n edge gateways with the highest comprehensive scores are selected as core gateways. Simultaneously, the data interaction protocol and collaborative working mechanism between the edge gateways and core gateways are determined. The calculation formula is as follows:

[0011] Si = λ*(Di / ∑Di) + μ*Wi + v*Ki / N

[0012] Where Si is the comprehensive score of the i-th edge gateway, Di is the comprehensive data collection volume of the i-th edge gateway, Wi is the concentration of the area where the i-th edge gateway is located, the closer to the center of the area, the higher Wi is, Ki is the gateway computing power load coefficient, N is the number of neighboring gateways, λ, μ and v are the calculation weights, and λ+μ+v=1.

[0013] The system presets task configurations for various scenarios, calls the corresponding task configuration based on the current application scenario, and generates multiple task allocation plans based on the task configuration. Using the weighted sum of energy consumption and time required to process the task as the objective function, energy efficiency optimization calculations are performed on each task allocation plan. The calculation formula is as follows:

[0014] E=∑ M k=1 (α*Ec+β*Em)

[0015] Where E is the true energy efficiency calculation index, Ec is the energy consumption cost, Em is the calculation efficiency, α and β are both calculation weights, α+β=1, and the plan with the smallest E is selected as the optimal task allocation plan.

[0016] Based on the energy efficiency optimization calculation results, the optimal allocation task plan that matches each core gateway is selected from multiple allocation task plans, and the core gateway schedules and executes the task according to the optimal allocation task plan.

[0017] Furthermore, the construction of the three-dimensional point cloud map involves first constructing a multi-source data fusion analysis model, simultaneously accessing meteorological data and infrared camera inspection images, and combining collected power generation data and equipment component status data to perform multi-source data fusion and feature extraction, thereby obtaining environmental features, component hot spot features, power generation features, and hardware status features, and establishing a three-dimensional point cloud map based on the obtained features.

[0018] Furthermore, the three-dimensional point cloud map is created by mapping feature vectors to three-dimensional space using a point cloud library, and establishing three-dimensional point cloud maps of component hot spot features, power generation features, and hardware status features under different environmental characteristics and service life conditions based on historical data.

[0019] Historical data includes data collected under normal and different fault conditions of new energy power generation equipment. The three-dimensional point cloud map contains corresponding point cloud areas for normal conditions and point cloud areas for different fault conditions.

[0020] Furthermore, the fault diagnosis process for the aforementioned new energy power generation equipment is as follows:

[0021] Based on the current service life and environmental characteristics of the target new energy power generation equipment, the current hot spot characteristics, power generation characteristics and hardware status characteristics of the components are input into the three-dimensional point cloud map. The Mahalanobis distance algorithm is used to calculate the difference between the current hot spot characteristics, power generation characteristics and hardware status characteristics of the components and each point cloud region. The point cloud region corresponding to the smallest difference value is selected to determine whether the target new energy power generation equipment is in a normal state.

[0022] The formula for calculating the difference value is as follows:

[0023] ΔS min =√[(F-ρR) T *∑ R -1 (F-ρR)]

[0024] Where, ΔS min Let F be the difference value, F be the current feature vector, ρR be the mean vector of the point cloud region R, and ∑ R Let be the covariance matrix of the point cloud region R;

[0025] If the target new energy power generation equipment is in a fault, the faulty equipment components are located based on the inspection images from the infrared camera, and the fault type of the target new energy power generation equipment is determined by combining the corresponding point cloud region.

[0026] Furthermore, the lifespan prediction of the new energy power generation equipment involves constructing an LSTM-based lifespan prediction model. Inputting the equipment's lifespan, fault type coding value, and minimum difference value, the model predicts the remaining lifespan of the faulty components and simultaneously generates a spare parts list and inspection work orders. The formula for calculating the remaining lifespan is as follows:

[0027] L rem =L t -L c -k*ΔS min *T err

[0028] Among them, L rem L represents the remaining service life of the failed component. t For the rated service life of the component, L c ΔS represents the usage time, k is the fault impact coefficient (different values ​​for different fault types), and ΔS is the fault duration. min For the minimum difference value, T err The duration of abnormal data collection.

[0029] Preferably, the strategy generation involves using a long short-term memory network to predict the output of the target new energy power generation equipment within a preset time period based on environmental characteristics, the power generation characteristics of each new energy power generation equipment, and hardware status characteristics. Based on the prediction results, a collaborative control strategy is generated for each new energy power generation equipment in the regional gateway network. After receiving the collaborative control strategy, the edge gateway adjusts the power operation of the new energy power generation equipment through a PID controller, shuts down or limits the power operation of faulty new energy power generation equipment, and improves the power operation of new energy power generation equipment in normal state, thereby maintaining the overall output stability of the power grid.

[0030] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the contents of a big data-based energy acquisition and monitoring system as described above.

[0031] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the contents of a big data-based energy harvesting and monitoring system as described above.

[0032] Compared to existing technologies, the beneficial effects of this application are as follows:

[0033] (1) This application combines a dynamic task allocation model with multi-index weighted scoring and energy efficiency optimization. It selects the core gateway based on real-time data volume, regional location concentration and gateway computing load, and generates a task plan with energy efficiency optimization as the goal. This achieves computing load balancing and data transmission latency minimization, realizes the reasonable selection of core gateways and balanced task allocation, and improves the overall computing power utilization and task execution efficiency of the system.

[0034] (2) This application constructs a three-dimensional point cloud map by fusing multi-source data and combines the Mahalanobis distance algorithm and LSTM model to perform fault judgment and remaining service life prediction. It can accurately identify fault types, locate faulty components, and generate a list of maintenance spare parts and work orders, thereby reducing false alarm rate and maintenance costs and extending equipment service life.

[0035] (3) Based on environmental characteristics and equipment status, this application uses a long short-term memory network to predict power generation output and generate a collaborative control strategy to dynamically adjust equipment power output, effectively shut down or limit faulty equipment, improve the operating efficiency of normal equipment, and thus maintain the overall output stability of the power grid. Attached Figure Description

[0036] In the attached diagram:

[0037] Figure 1 This is a schematic diagram of the system structure according to an embodiment of this application. Detailed Implementation

[0038] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of this application, but not all embodiments. Generally, the components of the embodiments of this application described and shown in the accompanying drawings can be arranged and designed in various different configurations. Example 1

[0039] like Figure 1 As shown, an energy acquisition and monitoring system based on big data includes:

[0040] The data acquisition module installs an edge gateway on each new energy power generation device within the power grid to collect power generation data and status data of device components;

[0041] The edge gateway has a built-in data compression algorithm to clean, cache, and compress data. It adaptively adjusts the sampling frequency and compression ratio based on the rate of change of the collected data and the quality of the communication link. When communication with the regional gateway network is interrupted, it caches critical data and prioritizes the transmission of abnormal data fragments after recovery.

[0042] The rate of change of the collected data is obtained by calculating the variance or derivative of the data sequence. When the rate of change is higher than the threshold, the sampling frequency is increased; when the rate of change is lower than the threshold, the sampling frequency is decreased.

[0043] The compression ratio is dynamically adjusted by measuring the signal-to-noise ratio (SNR) or bit error rate (BER): when the communication link quality is less than a threshold, the compression ratio is increased; when the communication link quality is greater than the threshold, the compression ratio is decreased.

[0044] When compressing data, high-frequency components are preserved first, and entropy coding is used for optimization.

[0045] Data caching and transmission utilize a built-in circular buffer in the edge gateway, employing a hierarchical caching strategy. Device fault data and critical operating parameters are classified as first-level cache, while ordinary operating data is classified as second-level cache. When the network is down, first-level cache data is saved first. After the network is restored, abnormal data fragments in the first-level cache are transmitted first through an abnormal data marker bit.

[0046] The algorithm identifies key operating parameters and marks them as high priority through mutation detection. These parameters are cached when the network is interrupted and transmitted preferentially after network recovery. The caching strategy is based on the LRU algorithm to ensure that critical data is not lost.

[0047] The task allocation module establishes a regional gateway network among all edge gateways, builds a dynamic task allocation model and deploys it to the edge gateways. It uses the dynamic task allocation model to select multiple core gateways from all edge gateways and assigns different task plans to different core gateways.

[0048] The task execution module enables the core gateway to perform point cloud map construction, fault diagnosis, lifespan prediction, and strategy generation according to the assigned task plan, thereby realizing the monitoring and management of new energy power generation equipment.

[0049] The task schedule for the core gateway includes:

[0050] A dynamic task allocation model is used to select core gateways based on the amount of data collected, regional location, and number of gateways. Different weights are assigned to the amount of data collected, regional location concentration, and gateway computing load as evaluation indicators. A comprehensive score is calculated for each edge gateway, and the top n edge gateways with the highest comprehensive scores are selected as core gateways. Simultaneously, the data interaction protocol and collaborative working mechanism between the edge gateways and core gateways are determined. The calculation formula is as follows:

[0051] Si = λ*(Di / ∑Di) + μ*Wi + v*Ki / N

[0052] Where Si is the comprehensive score of the i-th edge gateway, Di is the comprehensive data collection volume of the i-th edge gateway, Wi is the concentration of the area where the i-th edge gateway is located, the closer to the center of the area, the higher Wi is, Ki is the gateway computing power load coefficient, N is the number of neighboring gateways, λ, μ and v are the calculation weights, and λ+μ+v=1.

[0053] The system presets task configurations for various scenarios, calls the corresponding task configuration based on the current application scenario, and generates multiple task allocation plans based on the task configuration. Using the weighted sum of energy consumption and time required to process the task as the objective function, energy efficiency optimization calculations are performed on each task allocation plan. The calculation formula is as follows:

[0054] E=∑ M k=1 (α*Ec+β*Em)

[0055] Where E is the true energy efficiency calculation index, Ec is the energy consumption cost, Em is the calculation efficiency, α and β are both calculation weights, α+β=1, and the plan with the smallest E is selected as the optimal task allocation plan.

[0056] Based on the energy efficiency optimization calculation results, the optimal allocation task plan that matches each core gateway is selected from multiple allocation task plans, and the core gateway schedules and executes the task according to the optimal allocation task plan.

[0057] The dynamic task allocation model is based on reinforcement learning algorithms and deployed in edge gateways. Based on the amount of real-time collected data, regional location, and number of gateways, the model uses different weights of collected data volume (0.4), regional location concentration (0.3), and gateway computing load (0.3) as evaluation indicators to calculate the comprehensive score of each edge gateway. The model selects multiple core gateways by using the minimization of data transmission latency and the balancing of computing load as selection criteria.

[0058] The construction of a 3D point cloud map involves first building a multi-source data fusion analysis model, then simultaneously accessing meteorological data (irradiance, temperature, wind speed) and infrared camera inspection images, and combining collected power generation data (current, voltage, power) and equipment component status data to perform multi-source data fusion and feature extraction, thereby obtaining environmental features, component hot spot features, power generation features, and hardware status features, and then establishing a 3D point cloud map based on the obtained features.

[0059] 3D point cloud maps are created by mapping feature vectors to 3D space using a point cloud library. Based on historical data, they are used to create 3D point cloud maps of component hot spot features, power generation features, and hardware status features under different environmental characteristics and service life conditions.

[0060] Historical data includes data collected under normal and different fault conditions of new energy power generation equipment. The three-dimensional point cloud map contains corresponding point cloud areas for normal conditions and point cloud areas for different fault conditions.

[0061] The fault diagnosis process for new energy power generation equipment is as follows:

[0062] Based on the current service life and environmental characteristics of the target new energy power generation equipment, the current hot spot characteristics, power generation characteristics and hardware status characteristics of the components are input into the three-dimensional point cloud map. The Mahalanobis distance algorithm is used to calculate the difference between the current hot spot characteristics, power generation characteristics and hardware status characteristics of the components and each point cloud region. The point cloud region corresponding to the smallest difference value is selected to determine whether the target new energy power generation equipment is in a normal state.

[0063] The formula for calculating the difference value is as follows:

[0064] ΔS min =√[(F-ρR) T *∑ R -1 (F-ρR)]

[0065] Where, ΔS min Let F be the difference value, F be the current feature vector, ρR be the mean vector of the point cloud region R, and ∑ R Let be the covariance matrix of the point cloud region R;

[0066] If the target new energy power generation equipment is in a fault, the faulty equipment components are located based on the inspection images from the infrared camera, and the fault type of the target new energy power generation equipment is determined by combining the corresponding point cloud region.

[0067] Lifespan prediction for new energy power generation equipment involves constructing an LSTM-based lifespan prediction model. Inputting the equipment's lifespan, fault type coding value, and minimum difference value, the model predicts the remaining lifespan of faulty components and simultaneously generates a spare parts list and inspection work orders. The formula for calculating the remaining lifespan is as follows:

[0068] L rem =L t -L c -k*ΔS min *T err

[0069] Among them, L rem L represents the remaining service life of the failed component. t For the rated service life of the component, L c ΔS represents the usage time, k is the fault impact coefficient (different values ​​for different fault types), and ΔS is the fault duration. min For the minimum difference value, T err The duration of abnormal data collection.

[0070] Strategy generation involves using a long short-term memory network to predict the output of a target new energy power generation device within a preset time period, based on environmental characteristics, the power generation characteristics of each new energy power generation device, and hardware status characteristics. Based on the prediction results, a collaborative control strategy is generated for each new energy power generation device in the regional gateway network. After receiving the collaborative control strategy, the edge gateway adjusts the power operation of the new energy power generation device through a PID controller, shuts down or limits the power operation of faulty new energy power generation devices, and improves the power operation of new energy power generation devices in normal state, thereby maintaining the overall output stability of the power grid. Example 2

[0071] From a hardware perspective, this application provides an embodiment of an electronic device comprising all or part of a big data-based energy harvesting and monitoring system. The electronic device includes a service processor and a distributed memory. The service processor is connected to the memory, and the distributed memory stores a service self-management program configured to store machine-readable instructions. The service processor executes the service self-management program, and the instructions, when executed by the processor, implement the big data-based energy harvesting and monitoring system as described above. Example 3

[0072] The embodiments of this application also provide a computer-readable storage medium capable of implementing a big data-based energy harvesting and monitoring system with a server or client as the execution subject in the above embodiments. The computer-readable storage medium stores a computer program, which, when executed by a processor, implements all the contents of the big data-based energy harvesting and monitoring system with a server or client as the execution subject in the above embodiments.

[0073] The embodiments described above are merely preferred embodiments of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications, improvements, and substitutions without departing from the concept of this application, and these all fall within the protection scope of this application.

Claims

1. An energy acquisition and monitoring system based on big data, characterized in that, include: The data acquisition module is used to install an edge gateway on each new energy power generation device in the power grid to collect the power generation data and status data of the device components; The task allocation module is used to establish a regional gateway network among all edge gateways, build a dynamic task allocation model and deploy it to the edge gateways. Using the dynamic task allocation model, it selects multiple edge gateways as core gateways and assigns different task plans to different core gateways. The task plan allocation includes: A dynamic task allocation model is used to select core gateways based on the amount of data collected, regional location, and number of gateways. Different weights are assigned to the amount of data collected, regional location concentration, and gateway computing load as evaluation indicators. A comprehensive score is calculated for each edge gateway, and the top n edge gateways with the highest comprehensive scores are selected as core gateways. Simultaneously, the data interaction protocol and collaborative working mechanism between the edge gateways and core gateways are determined. The calculation formula is as follows: Si = λ*(Di / ∑Di) + μ*Wi + v*Ki / N Where Si is the comprehensive score of the i-th edge gateway, Di is the comprehensive data collection volume of the i-th edge gateway, Wi is the concentration of the area where the i-th edge gateway is located, the closer to the center of the area, the higher Wi is, Ki is the gateway computing power load coefficient, N is the number of neighboring gateways, λ, μ and v are the calculation weights, and λ+μ+v=1. The system presets task configurations for various scenarios, calls the corresponding task configuration based on the current application scenario, and generates multiple task allocation plans based on the task configuration. Using the weighted sum of energy consumption and time required to process the task as the objective function, energy efficiency optimization calculations are performed on each task allocation plan. The calculation formula is as follows: E=∑ M k=1 (α*Ec+β*Em) Where E is the true energy efficiency calculation index, Ec is the energy consumption cost, Em is the calculation efficiency, α and β are both calculation weights, α+β=1, and the plan with the smallest E is selected as the optimal task allocation plan. Based on the energy efficiency optimization calculation results, the optimal allocation task plan that matches each core gateway is selected from multiple sets of allocation task plans, and the core gateway schedules and executes the task according to the optimal allocation task plan. The task execution module enables the core gateway to perform point cloud map construction, fault diagnosis, lifespan prediction, and strategy generation according to the assigned task plan, thereby realizing the monitoring and management of new energy power generation equipment.

2. The energy acquisition and monitoring system based on big data according to claim 1, characterized in that, The edge gateway has a built-in data compression algorithm to clean, cache, and compress data. It adaptively adjusts the sampling frequency and compression ratio based on the rate of change of the collected data and the quality of the communication link. When communication with the regional gateway network is interrupted, it caches key data and prioritizes the transmission of abnormal data fragments after recovery.

3. The energy acquisition and monitoring system based on big data according to claim 1, characterized in that, The point cloud map construction involves first building a multi-source data fusion analysis model, simultaneously accessing meteorological data and infrared camera inspection images, and combining collected power generation data and equipment component status data to perform multi-source data fusion and feature extraction, obtaining environmental features, component hot spot features, power generation features, and hardware status features, and then establishing a three-dimensional point cloud map based on the obtained features.

4. The energy acquisition and monitoring system based on big data according to claim 3, characterized in that, The three-dimensional point cloud map is created by mapping feature vectors to three-dimensional space using a point cloud library, and based on historical data, it establishes three-dimensional point cloud maps of component hot spot features, power generation features, and hardware status features under different environmental characteristics and service life conditions. Historical data includes data collected under normal and different fault conditions of new energy power generation equipment. The three-dimensional point cloud map contains corresponding point cloud areas for normal conditions and point cloud areas for different fault conditions.

5. The energy acquisition and monitoring system based on big data according to claim 4, characterized in that, The fault diagnosis process for the new energy power generation equipment is as follows: Based on the current service life and environmental characteristics of the target new energy power generation equipment, the current hot spot characteristics, power generation characteristics and hardware status characteristics of the components are input into the three-dimensional point cloud map. The Mahalanobis distance algorithm is used to calculate the difference between the current hot spot characteristics, power generation characteristics and hardware status characteristics of the components and each point cloud region. The point cloud region corresponding to the smallest difference value is selected to determine whether the target new energy power generation equipment is in a normal state. The formula for calculating the difference value is as follows: ΔS min =√[(F-ρR) T *∑ R -1 (F-ρR)] Where, ΔS min Let F be the difference value, F be the current feature vector, ρR be the mean vector of the point cloud region R, and ∑ R Let be the covariance matrix of the point cloud region R; If the target new energy power generation equipment is in a fault, the faulty equipment components are located based on the inspection images from the infrared camera, and the fault type of the target new energy power generation equipment is determined by combining the corresponding point cloud region.

6. The energy acquisition and monitoring system based on big data according to claim 5, characterized in that, The lifespan prediction of the new energy power generation equipment involves constructing an LSTM-based lifespan prediction model. Inputting the equipment's lifespan, fault type coding value, and minimum difference value, the model predicts the remaining lifespan of the faulty components and simultaneously generates a spare parts list and inspection work orders. The formula for calculating the remaining lifespan is as follows: L rem =L t -L c -k*ΔS min *T err Among them, L rem L represents the remaining service life of the failed component. t For the rated service life of the component, L c ΔS represents the usage time, k is the fault impact coefficient (different values ​​for different fault types), and ΔS is the fault duration. min For the minimum difference value, T err The duration of abnormal data collection.

7. The energy acquisition and monitoring system based on big data according to claim 1, characterized in that, The strategy generation involves using a long short-term memory network to predict the output of the target new energy power generation equipment within a preset time period based on environmental characteristics, the power generation characteristics of each new energy power generation equipment, and hardware status characteristics. Based on the prediction results, a collaborative control strategy is generated for each new energy power generation equipment in the regional gateway network. After receiving the collaborative control strategy, the edge gateway adjusts the power operation of the new energy power generation equipment through a PID controller, shuts down or limits the power operation of faulty new energy power generation equipment, and improves the power operation of new energy power generation equipment in normal state, thereby maintaining the overall output stability of the power grid.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the content of the energy acquisition and monitoring system based on big data as described in claim 1.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the content of the energy acquisition and monitoring system based on big data as described in claim 1.

Images