Intelligent scheduling method for video equipment based on solar power supply

Abstract

The invention discloses an intelligent scheduling method for video equipment based on solar power supply, and relates to the technical field of energy management and video. Through multi-dimensional data acquisition, intelligent prediction and dynamic scheduling technologies, the energy management challenge faced by solar power supply video equipment in remote areas is effectively solved, the power generation amount, equipment power consumption and environmental change data are acquired in real time by using an Internet of Things sensor, and the energy management efficiency is improved by combining time sequence analysis and a machine learning algorithm. The power generation amount and power consumption fluctuation are accurately predicted, an energy supply and demand matching model is established based on a dynamic planning algorithm, a power distribution strategy is optimized, operation of high-priority equipment is guaranteed preferentially when the power generation amount is insufficient, meanwhile, the working mode of the equipment is dynamically adjusted through a reinforcement learning algorithm, and energy consumption optimization and self-adaptive matching are achieved. The energy utilization efficiency and the system stability are improved, the adaptive capacity to dynamic environment changes is enhanced, and continuous and stable operation of video equipment under complex conditions is ensured.

Description

Technical Field

[0001] The present invention relates to the technical fields of energy management and video technology, and specifically to an intelligent scheduling method for video devices powered by solar energy. Background Art

[0002] When deploying video devices powered by solar energy in remote areas, complex energy management challenges are faced. Due to the unstable local power grid, the devices can only rely on solar panels for power supply. The solar power generation is greatly affected by weather and seasonal changes, and the power consumption of video devices also varies dynamically according to usage conditions. To ensure the continuous and stable operation of the devices, it is a problem that needs to be solved currently to achieve intelligent scheduling by comprehensively considering the solar power generation, the power consumption characteristics of the devices, and the changing trends of environmental conditions. Summary of the Invention

[0003] The purpose of the present invention is to provide an intelligent scheduling method for video devices powered by solar energy, which effectively solves the energy management challenges faced by solar-powered video devices in remote areas.

[0004] The purpose of the present invention can be achieved through the following technical solutions:

[0005] The present application provides an intelligent scheduling method for video devices powered by solar energy, including the following steps:

[0006] An intelligent scheduling method for video devices powered by solar energy, including the following steps:

[0007] Obtain the real-time power generation data of the solar panels through Internet of Things sensors, and at the same time collect the operating status and power consumption data of the video devices, and combine the environmental change information to construct a multi-dimensional data set;

[0008] Use time series analysis algorithms to model the power generation fluctuations and power consumption dynamics, extract the periodic characteristics and trend changes in historical data, and generate short-term power generation and power consumption prediction results;

[0009] According to the prediction results, establish an energy supply-demand matching model, and use dynamic programming algorithms to calculate the optimal scheduling strategy. When the power generation is insufficient, give priority to allocating power resources to high-priority devices, and store the excess power in energy storage devices when the power generation is excessive;

[0010] Through the device power consumption characteristic database, obtain the power consumption curves and operating modes of each video device, combine the real-time device status and environmental change data, dynamically update the power consumption characteristic parameters, and then use reinforcement learning algorithms to optimize the scheduling strategy. Through historical operation data and real-time feedback information, adjust the device working modes and parameters to achieve adaptive matching of energy supply and demand;

[0011] When the energy supply is insufficient, according to the device priority and energy sharing mechanism, coordinate the energy distribution among multiple devices;

[0012] Through the edge computing node, real-time process the collected data and scheduling instructions, reduce the data transmission delay, and feedback the execution results of the scheduling strategy to the prediction model and optimization algorithm to continuously iterate and update the model parameters.

[0013] Furthermore, construct a multi-dimensional data set, specifically including:

[0014] By installing Internet of Things sensors on the solar panels, real-time collect the power generation data of the panels to obtain a data stream reflecting the power generation efficiency, then obtain the operation status data and power consumption data of the video devices, determine whether the devices are working properly by analyzing the operation status, and judge whether the energy consumption is abnormal according to the power consumption data;

[0015] Collect the environmental temperature and light intensity data, judge their impacts on the power generation efficiency of the solar panels and the operation of the video devices by analyzing the temperature and light change rules, and fuse the multi-source heterogeneous data of the solar panel power generation, video device operation status and power consumption, environmental temperature and light intensity to construct a multi-dimensional data set;

[0016] Perform data preprocessing on the multi-dimensional data set, use machine learning algorithms to establish a prediction model for the power generation efficiency of the solar panels, and predict the power generation according to environmental factors.

[0017] Furthermore, use time series analysis algorithms to model the power generation fluctuations and power consumption dynamics, extract the periodic characteristics and trend changes in the historical data, and generate short-term power generation and power consumption prediction results, specifically including:

[0018] Obtain the historical power generation and power consumption data, sort and preprocess the data according to the time series, remove the outliers and missing values, and obtain a normalized time series data set;

[0019] Use the seasonal decomposition algorithm to decompose the time series data set, split it into a trend term, a periodic term and a random term, and respectively extract the long-term trend change rules and periodic fluctuation characteristics of the power generation and power consumption data;

[0020] Use the support vector machine regression algorithm to model the decomposed trend term data, obtain a regression model describing the long-term change trends of the power generation and power consumption through training, and then use the Fourier transform algorithm to perform frequency domain analysis on the decomposed periodic term data to extract the main periodic components in the power generation and power consumption data, and obtain the frequency domain characteristics reflecting their periodic laws;

[0021] Combine the output result of the trend item regression model with the frequency domain characteristics of the periodic item, and use the weighted average method to fuse the prediction results of the two parts to obtain the predicted values of power generation and power consumption with periodic fluctuations;

[0022] Among them, the weight allocation for fusing the prediction results of the two parts is dynamically adjusted according to the prediction error or historical performance of the model.

[0023] Furthermore, after generating the short-term power generation and power consumption prediction results, it also includes: using the confidence interval estimation method to conduct uncertainty analysis on the predicted values of power generation and power consumption with periodic fluctuations to obtain the confidence interval of the predicted values, and then visually display the historical data, prediction results, and confidence intervals of power generation and power consumption to generate intuitive charts and reports.

[0024] Furthermore, according to the prediction results, establish an energy supply-demand matching model, and use the dynamic programming algorithm to calculate the optimal scheduling strategy, specifically including:

[0025] Obtain the power generation and power consumption data for a future period according to the prediction results, establish an energy supply-demand matching model, and then use the dynamic programming algorithm to divide the entire scheduling time into multiple stages. The state of each stage is determined by the power generation, power consumption, and power storage state at the current moment;

[0026] For each stage, by recursively solving sub-problems, calculate the optimal power distribution plan under different power generation and power consumption conditions. The goal is to minimize the power gap or surplus while meeting the electricity demand;

[0027] During the solution process, according to the device priority attribute, when the power generation is insufficient, give priority to allocating power resources to high-priority devices to ensure the normal operation of important devices. When the power generation is excessive, store the excess power in energy storage devices;

[0028] In the stage of dynamic programming, update the state of the next stage according to the current state and decision until the end point of the scheduling time is reached to obtain the optimal scheduling strategy for the entire time period.

[0029] Furthermore, after obtaining the optimal scheduling strategy, it also includes: converting the scheduling plan generated by the dynamic programming algorithm into specific control instructions to regulate the operation mode of video devices and the charge and discharge operations of energy storage devices. According to the power distribution plan, adjust the power state of video devices in real time, give priority to ensuring the stable operation of high-priority devices when the power generation is limited, and at the same time make efficient use of energy storage devices to store excess power when the power generation is excessive. During the operation process, continuously monitor the device state and the energy storage power, and dynamically optimize the scheduling strategy according to real-time data.

[0030] Further, through the device energy consumption characteristic database, obtain the power consumption curve and operating mode of each video device, and combine the real-time device status and environmental change data to dynamically update the energy consumption characteristic parameters, specifically including:

[0031] Based on the historical data in the device energy consumption characteristic database, establish a power consumption curve model and an operating mode model for each video device through machine learning algorithms, obtain the real-time status data and environmental change data of the video device, use them as input features, and make predictions through the power consumption curve model and the operating mode model to obtain the real-time energy consumption characteristic parameters of the device;

[0032] Based on the real-time energy consumption characteristic parameters of the device, determine whether the device is in the optimal operating state. If not, dynamically adjust the operating parameters of the device through an optimization algorithm to make it reach the optimal energy consumption state;

[0033] Feed back the optimized device operating parameters to the device control system, and perform real-time regulation on the device through the control system to make the device always in the optimal energy consumption state;

[0034] Continuously obtain the real-time status data and environmental change data of the device, dynamically update the power consumption curve model and the operating mode model through the incremental learning algorithm, so that the models can adapt to the changes of the device and the environment, and then regularly compare and analyze the actual energy consumption data and predicted energy consumption data of the device, and fine-tune the models through the error backpropagation algorithm;

[0035] Store the optimized device energy consumption characteristic parameters in the device energy consumption characteristic database as historical data for model training and optimization, forming a closed-loop feedback mechanism.

[0036] Further, adopt a reinforcement learning algorithm to optimize the scheduling strategy. Through historical operation data and real-time feedback information, adjust the device working mode and parameters to perform adaptive matching of energy supply and demand, specifically including: obtain the historical operation data and real-time feedback information of the device, dynamically update the device energy consumption characteristic parameters to obtain the latest energy consumption model of the device, and according to the updated energy consumption model, adopt a reinforcement learning algorithm to optimize the device scheduling strategy, and perform adaptive matching of energy supply and demand by adjusting the device working mode and operating parameters;

[0037] When optimizing the scheduling strategy, judge whether the current energy supply meets the device operation requirements. When the energy supply is sufficient, execute the device scheduling according to the optimized scheduling strategy. When it is judged that the energy supply is insufficient to meet the operation requirements of all devices, determine the energy distribution weights of each device according to the preset device priorities;

[0038] Obtain the energy sharing mechanism and protocol between devices. Through the energy sharing mechanism, allocate limited energy among multiple devices, and prioritize ensuring the energy supply of high-priority devices. According to the device priority and energy distribution weight, use an intelligent optimization algorithm to solve the optimal energy distribution scheme among multiple devices and generate the energy distribution strategy between devices.

[0039] Send the optimized energy distribution strategy between devices to each device. Through the energy coordination and sharing between devices, ensure the normal operation of key devices when the energy supply is insufficient.

[0040] Furthermore, the edge computing node is used to process the collected data and scheduling instructions in real time, specifically including:

[0041] Obtain the computing power index and network status index of the edge computing node. Then, according to the obtained computing power index and network status index, combined with the preset threshold range, judge the load situation and network quality situation of the edge computing node. When the load of the edge computing node is too high or the network quality is poor, dynamically reduce the frequency and granularity of data collection.

[0042] Establish a mapping model between the load and network quality of the edge computing node and the data collection frequency and granularity through a machine learning algorithm. According to the established mapping model, dynamically adjust the configuration parameters of the data collection task to adapt to the load change and network fluctuation of the edge computing node.

[0043] Furthermore, after processing the collected data and scheduling instructions in real time through the edge computing node, it also includes: using an online learning algorithm to update the parameters of the prediction model in real time according to the execution effect of the scheduling strategy. Through a reinforcement learning algorithm, autonomously learn and optimize the scheduling strategy according to the historical scheduling data and the current system state, and then construct a multi-level scheduling decision framework. Dynamically adjust the scheduling strategy and execution plan according to the urgency of the task and the availability of resources. Use a federated learning algorithm to jointly train the prediction model using the data of multiple edge nodes.

[0044] Model the topological relationship and communication mode between edge computing nodes through a graph neural network algorithm, optimize the data transmission path and scheduling strategy, and according to the dependency relationship and priority of edge computing tasks, use a DAG scheduling algorithm to generate the optimal task execution order and resource allocation scheme to minimize the task completion time.

[0045] The beneficial effects of the present invention are:

[0046] Through multi-dimensional data collection and intelligent prediction algorithms, the present invention effectively solves the energy management problem of solar-powered video communication devices in remote areas under complex environments. By using Internet of Things sensors to collect power generation, device power consumption, and environmental change data in real time, and combining time series analysis and machine learning models, it accurately predicts short-term fluctuations in power generation and power consumption. Then, a dynamic programming algorithm is used to establish an energy supply-demand matching model and optimize the power distribution strategy to ensure that high-priority devices are given priority when power generation is insufficient, and excess power is stored in energy storage devices when power generation is excessive. This significantly improves energy utilization efficiency, enhances the stability and reliability of the system in a dynamic environment, and effectively solves the problem of power generation fluctuations caused by weather and seasonal changes;

[0047] Through the device energy consumption characteristic database and reinforcement learning algorithm, the present invention realizes the energy consumption optimization and adaptive scheduling of video communication devices. It dynamically updates the energy consumption characteristic parameters according to the real-time device status and environmental changes, and uses the reinforcement learning algorithm to optimize the scheduling strategy and adjust the device working mode and parameters to achieve the best match between energy supply and demand. When energy is insufficient, the energy distribution among multiple devices is coordinated through the device priority and energy sharing mechanism to ensure the continuous operation of key devices. This not only optimizes the device energy consumption but also improves the adaptability to changes in the device operating state, solving the management problem caused by the dynamic change of video communication device power consumption;

[0048] With the synergistic effect of edge computing nodes and various advanced algorithms, the present invention significantly improves the system response speed and overall performance. By real-time processing of the collected data and scheduling instructions, the edge computing nodes reduce the data transmission delay and dynamically adjust the data collection frequency and granularity according to the computing power and network conditions. Combining technologies such as online learning, reinforcement learning, federated learning, and graph neural networks, it can quickly adapt to environmental changes, optimize the data transmission path and scheduling strategy. In addition, the DAG scheduling algorithm further optimizes the task execution order and resource allocation, minimizing the task completion time. This not only improves the real-time performance and robustness of the system but also enhances its adaptability in complex network environments, solving the scheduling delay problem caused by unstable networks in remote areas and providing strong support for the efficient operation of video communication devices. Brief Description of the Drawings

[0049] For better understanding and implementation, the technical solutions of this application will be described in detail below with reference to the accompanying drawings.

[0050] Figure 1 It is a schematic flowchart of an intelligent scheduling method for video communication devices powered by solar energy provided by this application;

[0051] Figure 2 It is a schematic flowchart of constructing a multi-dimensional data set for an intelligent scheduling method for video communication devices powered by solar energy provided by this application;

[0052] Figure 3 A schematic flow chart for generating short-term power generation and power consumption prediction results for an intelligent scheduling method of a video device powered by solar energy provided in this application. Detailed implementation manners

[0053] To further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, exemplary embodiments will be described in detail herein, and the examples are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all implementation manners consistent with this application. On the contrary, they are merely examples of methods and systems consistent with some aspects of this application as detailed in the appended claims.

[0054] The terms used in this application are only for the purpose of describing specific embodiments and are not intended to limit this application. The singular forms "a", "the", and "said" used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0055] The following will, in conjunction with the accompanying drawings and preferred embodiments, elaborate in detail on the specific implementation manners, features, and effects of the present invention.

[0056] Please refer to Figures 1 - 3 , this embodiment provides an intelligent scheduling method for a video device powered by solar energy, including the following steps:

[0057] S1. Obtain real-time power generation data of the solar panel through Internet of Things sensors, and at the same time collect the operating status and power consumption data of the video device, and combine environmental change information such as ambient temperature and light intensity to construct a multi-dimensional data set;

[0058] Furthermore, constructing a multi-dimensional data set specifically includes:

[0059] S11. Install Internet of Things sensors on the solar panel to collect the power generation data of the panel in real time, obtain a data stream reflecting the power generation efficiency, then obtain the operating status data and power consumption data of the video device, determine whether the device is working properly by analyzing the operating status, and judge whether the energy consumption is abnormal according to the power consumption data;

[0060] S12. Collect data on environmental temperature and light intensity. By analyzing the variation patterns of temperature and light, determine their impacts on the power generation efficiency of solar panels and the operation of video devices. Integrate multi-source heterogeneous data such as the power generation of solar panels, the operation status and power consumption of video devices, environmental temperature, and light intensity to construct a multi-dimensional data set.

[0061] S13. Perform data preprocessing on the multi-dimensional data set, including data cleaning, data normalization, etc., to improve data quality. Use machine learning algorithms such as support vector machines or random forests to establish a prediction model for the power generation efficiency of solar panels and predict the power generation based on environmental factors.

[0062] Based on the correlation analysis method, study the internal relationships between environmental temperature, light intensity, the power generation efficiency of solar panels, and the energy consumption of video devices, and optimize the system energy-saving scheme.

[0063] Specifically, by deploying Internet of Things sensors on solar panels and video devices, multi-dimensional data such as power generation, device operation status, power consumption, environmental temperature, and light intensity are collected in real time, and data fusion and preprocessing are performed. This method can effectively solve the energy management problems faced by solar-powered video devices in remote areas. By establishing a power generation efficiency prediction model and correlation analysis, the power generation is accurately predicted and the energy consumption allocation is optimized, thereby improving energy utilization efficiency, ensuring the stable operation of video devices under complex environmental conditions, and at the same time reducing the risk of service interruption caused by energy shortage or device failure. Through data preprocessing and modeling, high-quality data support is provided for energy supply and demand prediction and scheduling.

[0064] S2. Use time series analysis algorithms to model the power generation fluctuations and power consumption dynamics, extract the periodic characteristics and trend changes in historical data, and generate short-term power generation and power consumption prediction results.

[0065] Furthermore, use time series analysis algorithms to model the power generation fluctuations and power consumption dynamics, extract the periodic characteristics and trend changes in historical data, and generate short-term power generation and power consumption prediction results, specifically including:

[0066] S21. Obtain historical power generation and power consumption data, sort and preprocess the data according to the time series, remove outliers and missing values, and obtain a normalized time series data set.

[0067] S22. Use the seasonal decomposition algorithm to decompose the time series data set, split it into a trend term, a periodic term, and a random term, and respectively extract the long-term trend change laws and periodic fluctuation characteristics of the power generation and power consumption data.

[0068] S23. Use the support vector machine regression algorithm to model the decomposed trend term data, obtain a regression model describing the long-term change trends of power generation and power consumption through training, and then use the Fourier transform algorithm to perform frequency domain analysis on the decomposed periodic term data, extract the main periodic components in the power generation and power consumption data, and obtain the frequency domain characteristics reflecting their periodic laws;

[0069] S24. Combine the output results of the trend term regression model with the frequency domain characteristics of the periodic term, and use the weighted average method to fuse the prediction results of the two parts to obtain the predicted values of power generation and power consumption with periodic fluctuations;

[0070] Among them, the weight allocation for fusing the prediction results of the two parts is dynamically adjusted according to the prediction error or historical performance of the model.

[0071] Further, after generating the short-term power generation and power consumption prediction results, it also includes: using the confidence interval estimation method (such as Monte Carlo simulation or bootstrap method) to perform uncertainty analysis on the predicted values of power generation and power consumption with periodic fluctuations, obtain the confidence interval of the predicted values, and then visually display information such as the historical data, prediction results, and confidence interval of power generation and power consumption to generate intuitive charts and reports, for example, in the form of time series charts, confidence interval charts, or heat maps, to facilitate analysts to grasp the operating status and future trends of the power generation system and provide a decision-making basis for production scheduling and optimal control.

[0072] Specifically, by using the time series analysis algorithm to model the solar power generation and the power consumption of video devices, extract the periodic characteristics and trend changes in the historical data, and generate short-term prediction results, this method can effectively solve the energy management problems faced by solar-powered video devices in remote areas. Specifically, by sorting, preprocessing, seasonal decomposition, and extracting trend and cycle characteristics from the historical data, combining the support vector machine regression and Fourier transform algorithms, generate high-precision predicted values of power generation and power consumption; in addition, perform uncertainty analysis on the prediction results through the confidence interval estimation method, and visually display the historical data, prediction results, and confidence interval, providing a reliable basis for scheduling decisions. This method not only improves the prediction accuracy but also enhances the adaptability and robustness of the scheduling strategy, ensuring the efficient utilization of energy and the stable operation of equipment under complex environmental conditions. Through accurate prediction of power generation and power consumption, provide a decision-making basis for energy supply-demand matching and scheduling strategies.

[0073] S3. According to the prediction results, establish an energy supply-demand matching model, use the dynamic programming algorithm to calculate the optimal scheduling strategy, preferentially allocate power resources to high-priority devices when the power generation is insufficient, and store the excess power in the energy storage device when the power generation is excessive;

[0074] Furthermore, based on the prediction results, an energy supply-demand matching model is established, and the dynamic programming algorithm is used to calculate the optimal scheduling strategy, which specifically includes:

[0075] According to the prediction results, the power generation and power consumption data for a period of time in the future are obtained, and an energy supply-demand matching model is established. The model includes attributes such as power generation, power consumption, power distribution, and priority. Then, the dynamic programming algorithm is used to divide the entire scheduling time into multiple stages. The state of each stage is determined by the power generation, power consumption, and power storage state at the current moment;

[0076] For each stage, by recursively solving sub-problems, the optimal power distribution scheme under different power generation and power consumption conditions is calculated. The goal is to minimize the power gap or surplus while meeting the electricity demand;

[0077] During the solution process, according to the device priority attribute, when the power generation is insufficient, the power resources are preferentially allocated to high-priority devices to ensure the normal operation of important devices. When the power generation is excessive, the excess power is stored in energy storage devices such as batteries or supercapacitors for subsequent use;

[0078] In the stage of dynamic programming, according to the current state and decision, the state of the next stage is updated until the end of the scheduling time, and the optimal scheduling strategy for the entire time period is obtained.

[0079] Furthermore, after obtaining the optimal scheduling strategy, it also includes: converting the scheduling scheme generated by the dynamic programming algorithm into specific control instructions to accurately regulate the operation mode of the video device and the charge and discharge operations of the energy storage device. According to the power distribution scheme, the power state of the video device is adjusted in real time to ensure the stable operation of high-priority devices with limited power generation. At the same time, when the power generation is excessive, the energy storage device is efficiently used to store the excess power. During the operation process, the device state and the energy storage power are continuously monitored, and the scheduling strategy is dynamically optimized according to the real-time data to adapt to the changes in the device operation characteristics and environmental conditions.

[0080] Specifically, by establishing an energy supply-demand matching model and using the dynamic programming algorithm to calculate the optimal scheduling strategy, this method can effectively solve the energy management problem of solar-powered video devices in remote areas under the conditions of fluctuating power generation and dynamic changes in power consumption. Based on the prediction results of power generation and power consumption, the model recursively solves the optimal power distribution plan in stages. When the power generation is insufficient, it gives priority to ensuring the operation of high-priority devices. When the power generation is excessive, the excess power is stored in energy storage devices, thereby minimizing the power gap or excess. Then, the scheduling plan is converted into specific control instructions to adjust the operation mode of the devices and the charging and discharging operations of the energy storage devices in real time, and the scheduling strategy is dynamically optimized according to the real-time monitoring data to adapt to the changes in the device operation characteristics and environmental conditions. This method not only improves the energy utilization efficiency but also enhances the stability and adaptability of the system, ensuring the continuous and stable operation of video devices in complex environments. By optimizing the scheduling strategy through the dynamic programming algorithm, it ensures the efficient utilization of energy under the condition of fluctuating power generation.

[0081] S4. Obtain the power consumption curve and operation mode of each video device through the device energy consumption characteristic database. Combine the real-time device status and environmental change data to dynamically update the energy consumption characteristic parameters. Then, use the reinforcement learning algorithm to optimize the scheduling strategy. Through historical operation data and real-time feedback information, adjust the device working mode and parameters to achieve an adaptive matching of energy supply and demand.

[0082] When the energy supply is insufficient, according to the device priority and energy sharing mechanism, coordinate the energy distribution among multiple devices to ensure the continuous operation of key devices.

[0083] Furthermore, obtain the power consumption curve and operation mode of each video device through the device energy consumption characteristic database. Combine the real-time device status and environmental change data to dynamically update the energy consumption characteristic parameters, specifically including:

[0084] Based on the historical data in the device energy consumption characteristic database, establish a power consumption curve model and an operation mode model for each video device through machine learning algorithms. Obtain the real-time status data and environmental change data of the video device, use them as input features, and predict through the power consumption curve model and the operation mode model to obtain the real-time energy consumption characteristic parameters of the device.

[0085] According to the real-time energy consumption characteristic parameters of the device, determine whether the device is in the best operation state. If not, dynamically adjust the operation parameters of the device through an optimization algorithm to make it reach the optimal energy consumption state.

[0086] Feed the optimized device operation parameters back to the device control system, and perform real-time regulation on the device through the control system to ensure that the device is always in the best energy consumption state.

[0087] Continuously obtain the real-time status data of the device and the environmental change data, and dynamically update the power consumption curve model and the operation mode model through the incremental learning algorithm, so that the model can adapt to the changes of the device and the environment. Then, regularly compare and analyze the actual energy consumption data and the predicted energy consumption data of the device, and fine-tune the model through the error backpropagation algorithm to improve the prediction accuracy of the model;

[0088] Store the optimized device energy consumption characteristic parameters in the device energy consumption characteristic database as historical data for subsequent model training and optimization, forming a closed-loop feedback mechanism to continuously improve the energy-saving effect.

[0089] Furthermore, adopt the reinforcement learning algorithm to optimize the scheduling strategy. Through the historical operation data and real-time feedback information, adjust the device working mode and parameters to achieve the adaptive matching of energy supply and demand. Specifically, it includes: obtaining the historical operation data and real-time feedback information of the device, dynamically updating the device energy consumption characteristic parameters to obtain the latest energy consumption model of the device. According to the updated energy consumption model, adopt the reinforcement learning algorithm to optimize the device scheduling strategy, and realize the adaptive matching of energy supply and demand by adjusting the device working mode and operation parameters;

[0090] When optimizing the scheduling strategy, judge whether the current energy supply meets the device operation requirements. When the energy supply is sufficient, execute the device scheduling according to the optimized scheduling strategy. When it is judged that the energy supply is insufficient to meet the operation requirements of all devices, determine the energy allocation weight of each device according to the preset device priority;

[0091] Obtain the energy sharing mechanism and protocol among devices, and distribute the limited energy among multiple devices through the energy sharing mechanism, giving priority to ensuring the energy supply of high-priority devices. According to the device priority and energy allocation weight, adopt an intelligent optimization algorithm, such as the particle swarm optimization algorithm, to solve the optimal energy allocation scheme among multiple devices and generate the energy allocation strategy among devices;

[0092] Send the optimized energy allocation strategy among devices to each device. Through the energy coordination and sharing among devices, when the energy supply is insufficient, it can still ensure the normal operation of key devices and improve the overall energy utilization efficiency.

[0093] Specifically, by dynamically updating the energy consumption characteristic parameters of video devices and using reinforcement learning to optimize the scheduling strategy, this method solves the problem of mismatch between the dynamic changes in device energy consumption and energy supply in solar power systems in remote areas. Specifically, by combining the device energy consumption characteristic database and real-time data, a machine learning model is used to dynamically adjust the device operation parameters to ensure that the device is in the optimal energy consumption state. At the same time, through the reinforcement learning algorithm, the scheduling strategy is optimized based on historical data and real-time feedback. When the energy is insufficient, the operation of critical devices is prioritized, and the energy sharing mechanism is used to coordinate the energy distribution among multiple devices. This method not only improves the energy utilization efficiency but also enhances the system's adaptability and robustness, ensuring the stable operation of video devices in complex environments. By dynamically updating the energy consumption parameters and optimizing the scheduling strategy through reinforcement learning, the system's adaptive ability and energy-saving effect are further improved.

[0094] S5. The edge computing node processes the collected data and scheduling instructions in real time to reduce data transmission latency, improve the response speed and execution efficiency of scheduling decisions, and feedback the execution results of the scheduling strategy to the prediction model and optimization algorithm to continuously iteratively update the model parameters and improve the prediction accuracy and robustness of the scheduling strategy.

[0095] Furthermore, the edge computing node processes the collected data and scheduling instructions in real time, specifically including:

[0096] Obtain the computing power metrics and network status metrics of the edge computing node, including parameters such as CPU occupancy, memory occupancy, network bandwidth, and network latency. Then, based on the obtained computing power metrics and network status metrics and in combination with the preset threshold range, judge the load situation and network quality of the edge computing node. When the load of the edge computing node is too high or the network quality is poor, the data collection frequency and granularity are dynamically reduced to reduce the occupation of computing and network resources by data collection;

[0097] Through machine learning algorithms such as decision trees and support vector machines, establish a mapping model between the load and network quality of the edge computing node and the data collection frequency and granularity. According to the established mapping model, dynamically adjust the configuration parameters of the data collection task, including sampling interval, data compression ratio, transmission protocol, etc., to adapt to the load changes and network fluctuations of the edge computing node.

[0098] Furthermore, after the edge computing node processes the collected data and scheduling instructions in real time, it also includes: adopting an online learning algorithm to update the parameters of the prediction model in real time according to the execution effect of the scheduling strategy to improve the prediction accuracy and adaptability. For example, using online learning algorithms such as Hoeffding trees can process streaming data and update the model in real time to ensure that the model can quickly adapt to environmental changes.

[0099] Through the reinforcement learning algorithm, based on historical scheduling data and the current system state, it autonomously learns and optimizes the scheduling strategy to improve the efficiency and robustness of scheduling. Among them, reinforcement learning can effectively handle optimization problems with nonlinearity and randomness, and is particularly suitable for resource scheduling in dynamic environments. A multi-level scheduling decision framework is constructed. According to the urgency of tasks and the availability of resources, the scheduling strategy and execution plan are dynamically adjusted. This framework can better cope with the heterogeneity and geographical dispersion of resources in the edge computing environment. Then, the federated learning algorithm is adopted to utilize the data of multiple edge nodes to jointly train the prediction model and improve the generalization ability of the model. Federated learning can integrate the data characteristics of different nodes while protecting data privacy and enhance the adaptability of the model.

[0100] Through the graph neural network algorithm, the topological relationship and communication mode between edge computing nodes are modeled to optimize the data transmission path and scheduling strategy. According to the dependency relationship and priority of edge computing tasks, the DAG scheduling algorithm is adopted to generate the optimal task execution order and resource allocation scheme to minimize the task completion time.

[0101] Specifically, through the edge computing nodes to process the collected data and scheduling instructions in real time, combined with the synergistic effects of technologies such as reinforcement learning, federated learning, and graph neural network, this method effectively solves the problem of efficient scheduling of solar-powered video communication devices in remote areas in a dynamic environment. Specifically, using reinforcement learning to optimize the scheduling strategy can dynamically adjust the device operation mode according to real-time feedback to ensure that key devices are preferentially guaranteed to operate when the energy is insufficient. Federated learning, on the other hand, integrates the data of multiple edge nodes to jointly train the prediction model, improves the generalization ability of the model, and protects data privacy at the same time. The graph neural network further optimizes the data transmission path and scheduling strategy, and ensures the efficient execution of tasks by modeling the topological relationship between nodes. In addition, the DAG scheduling algorithm combines the task dependency relationship and priority to generate the optimal task execution order, significantly improving the task scheduling efficiency. The synergistic effects of these technologies not only improve the response speed and prediction accuracy of the system, but also enhance the overall performance and robustness, and are applicable to practical scenarios such as remote area monitoring and distributed energy management. Through the synergistic effects of edge computing and multiple machine learning algorithms, the overall performance and robustness of the system are improved.

[0102] As mentioned above, it is only a preferred embodiment of the present invention and does not impose any form of limitation on the present invention. Although the present invention has been disclosed above with a preferred embodiment, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or modifications into equivalent embodiments with equivalent changes within the scope of the technical solution of the present invention. However, as long as it does not depart from the content of the technical solution of the present invention, any brief modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. A method for intelligent scheduling of video equipment based on solar power supply, characterized in that: The steps include: The real-time power generation data of solar panels is obtained through IoT sensors, while the operating status and power consumption data of video equipment are collected, combined with environmental change information to build a multidimensional data set; Use time series analysis algorithms to model power generation fluctuations and power consumption dynamics, extract periodic features and trend changes in historical data, and generate short-term power generation and power consumption forecast results; Based on the prediction results, an energy supply and demand matching model is established, and the optimal scheduling strategy is calculated using a dynamic programming algorithm. When power generation is insufficient, power resources are allocated to high-priority equipment first, and when power generation is excessive, excess power is stored in energy storage equipment. The power consumption curve and operation mode of each video equipment are obtained through the equipment energy consumption characteristic database. The energy consumption characteristic parameters are dynamically updated in combination with the real-time equipment status and environmental change data. The scheduling strategy is then optimized using the reinforcement learning algorithm. The equipment working mode and parameters are adjusted through historical operation data and real-time feedback information to achieve adaptive matching of energy supply and demand. When the energy supply is insufficient, the energy distribution among multiple devices is coordinated according to the device priority and energy sharing mechanism; The edge computing nodes are used to process collected data and scheduling instructions in real time, reduce data transmission delays, and feed back the execution results of the scheduling strategy to the prediction model and optimization algorithm, continuously iterating and updating the model parameters.

2. The method for intelligent scheduling of video equipment based on solar power supply according to claim 1, characterized in that: Constructing a multidimensional dataset, including: By installing IoT sensors on solar panels, the power generation data of the panels can be collected in real time to obtain a data stream reflecting the power generation efficiency. Then, the operating status data and power consumption data of the video equipment can be obtained. By analyzing the operating status, it can be determined whether the equipment is working properly, and whether the energy consumption is abnormal based on the power consumption data. Collect ambient temperature and light intensity data, analyze the temperature and light change patterns, determine their impact on the power generation efficiency of solar panels and the operation of video equipment, fuse the multi-source heterogeneous data of solar panel power generation, video equipment operation status and power consumption, ambient temperature and light intensity, and construct a multi-dimensional data set; Data preprocessing is performed on multidimensional data sets, and a machine learning algorithm is used to establish a solar panel power generation efficiency prediction model to predict power generation based on environmental factors.

3. The method for intelligent scheduling of video equipment based on solar power supply according to claim 1, characterized in that: The time series analysis algorithm is used to model the power generation fluctuation and power consumption dynamics, extract the periodic characteristics and trend changes in historical data, and generate short-term power generation and power consumption forecast results, including: Obtain historical power generation and power consumption data, sort and preprocess the data according to the time series, remove outliers and missing values, and obtain a normalized time series data set; The seasonal decomposition algorithm is used to decompose the time series data set into trend terms, periodic terms and random terms, respectively extracting the long-term trend change law and periodic fluctuation characteristics of power generation and power consumption data; The support vector machine regression algorithm is used to model the decomposed trend item data. Through training, a regression model describing the long-term trend of power generation and power consumption is obtained. Then, the Fourier transform algorithm is used to perform frequency domain analysis on the decomposed periodic item data, extract the main periodic components in the power generation and power consumption data, and obtain the frequency domain characteristics reflecting their periodic laws. The output results of the trend term regression model are combined with the frequency domain characteristics of the periodic term, and the two parts of the prediction results are fused using the weighted average method to obtain the periodically fluctuating power generation and power consumption prediction values; Among them, the weight distribution for fusing the two parts of the prediction results is dynamically adjusted according to the prediction error or historical performance of the model.

4. The method for intelligent scheduling of video equipment based on solar power supply according to claim 3 is characterized in that: After generating the short-term power generation and power consumption forecast results, it also includes: using the confidence interval estimation method to perform uncertainty analysis on the periodically fluctuating power generation and power consumption forecast values, obtaining the confidence interval of the forecast value, and then visually displaying the historical data of power generation and power consumption, forecast results and confidence interval information to generate intuitive charts and reports.

5. The method for intelligent scheduling of video equipment based on solar power supply according to claim 1, characterized in that: According to the prediction results, an energy supply and demand matching model is established, and the dynamic programming algorithm is used to calculate the optimal scheduling strategy, which includes: According to the prediction results, the power generation and power consumption data in the future are obtained, and the energy supply and demand matching model is established. Then, the dynamic programming algorithm is used to divide the entire scheduling time into multiple stages. The state of each stage is determined by the power generation, power consumption and power storage state at the current moment; For each stage, the optimal power allocation scheme under different power generation and power consumption conditions is calculated by recursively solving sub-problems. The goal is to minimize the power gap or surplus while meeting the power demand. During the solution process, according to the equipment priority attributes, when the power generation is insufficient, the power resources are allocated to the high-priority equipment first to ensure the normal operation of important equipment. When the power generation is excessive, the excess power is stored in the energy storage device. In the dynamic programming stage, the state of the next stage is updated according to the current state and decision until the end of the scheduling time is reached, and the optimal scheduling strategy for the entire time period is obtained.

6. The method for intelligent scheduling of video equipment based on solar power supply according to claim 5 is characterized in that: After obtaining the optimal scheduling strategy, it also includes: converting the scheduling plan generated by the dynamic programming algorithm into specific control instructions, regulating the operating mode of the video equipment and the charging and discharging operations of the energy storage equipment, adjusting the power state of the video equipment in real time according to the power distribution plan, giving priority to ensuring the stable operation of high-priority equipment when power generation is limited, and efficiently using energy storage equipment to store excess power when power generation is in excess. During operation, the equipment status and energy storage power are continuously monitored, and the scheduling strategy is dynamically optimized based on real-time data.

7. The method for intelligent scheduling of video equipment based on solar power supply according to claim 1, characterized in that: The power consumption curve and operation mode of each video equipment are obtained through the equipment energy consumption characteristic database. The energy consumption characteristic parameters are dynamically updated in combination with the real-time equipment status and environmental change data. Specifically, the following are included: Based on the historical data in the equipment energy consumption characteristic database, the power consumption curve model and operation mode model of each video equipment are established through machine learning algorithms, and the real-time status data and environmental change data of the video equipment are obtained. The data are used as input features, and the power consumption curve model and operation mode model are used for prediction to obtain the real-time energy consumption characteristic parameters of the equipment. According to the real-time energy consumption characteristic parameters of the equipment, determine whether the equipment is in the best operating state. If not, dynamically adjust the operating parameters of the equipment through the optimization algorithm to achieve the optimal energy consumption state; Feedback the optimized equipment operating parameters to the equipment control system, and use the control system to adjust the equipment in real time to keep the equipment in the best energy consumption state; Continuously obtain the real-time status data of the device and the environmental change data, dynamically update the power consumption curve model and the operation mode model through the incremental learning algorithm, so that the model can adapt to the changes of the device and the environment, and then regularly compare and analyze the actual energy consumption data and the predicted energy consumption data of the device, and fine-tune the model through the error back propagation algorithm; The optimized equipment energy consumption characteristic parameters are stored in the equipment energy consumption characteristic database and used as historical data for model training and optimization to form a closed-loop feedback mechanism.

8. The method for intelligent scheduling of video equipment based on solar power supply according to claim 1, characterized in that: Adopt reinforcement learning algorithm to optimize scheduling strategy, adjust equipment working mode and parameters through historical operation data and real-time feedback information, and perform adaptive matching between energy supply and demand. Specifically, it includes: obtaining historical operation data and real-time feedback information of equipment, dynamically updating equipment energy consumption characteristic parameters, obtaining the latest energy consumption model of equipment, and adopting reinforcement learning algorithm to optimize equipment scheduling strategy according to the updated energy consumption model, and adaptive matching between energy supply and demand by adjusting equipment working mode and operation parameters; When optimizing the scheduling strategy, determine whether the current energy supply meets the equipment operation requirements. If the energy supply is sufficient, perform equipment scheduling according to the optimized scheduling strategy. If it is determined that the energy supply is insufficient to meet the operation requirements of all equipment, determine the energy allocation weight of each equipment according to the preset equipment priority. Obtain the energy sharing mechanism and protocol between devices, distribute limited energy among multiple devices through the energy sharing mechanism, give priority to the energy supply of high-priority devices, and use intelligent optimization algorithms to solve the optimal energy allocation plan among multiple devices based on device priority and energy allocation weight, and generate energy allocation strategies among devices; The optimized energy allocation strategy between devices is sent to each device. Through energy coordination and sharing between devices, the normal operation of key equipment is guaranteed when the energy supply is insufficient.

9. The method for intelligent scheduling of video equipment based on solar power supply according to claim 1, characterized in that: The edge computing nodes process the collected data and dispatch instructions in real time, including: Obtain the computing power indicators and network status indicators of the edge computing nodes, and then judge the load and network quality of the edge computing nodes based on the obtained computing power indicators and network status indicators and the preset threshold range; when the edge computing node load is too high or the network quality is poor, dynamically reduce the frequency and granularity of data collection; A mapping model between the edge computing node load and network quality and the data collection frequency and granularity is established through machine learning algorithms. According to the established mapping model, the configuration parameters of the data collection task are dynamically adjusted to adapt to the load changes and network fluctuations of the edge computing nodes.

10. The method for intelligent scheduling of video equipment based on solar power supply according to claim 9, characterized in that: After edge computing nodes process collected data and scheduling instructions in real time, the following also include: using online learning algorithms to update the parameters of the prediction model in real time according to the execution effect of the scheduling strategy; using reinforcement learning algorithms to autonomously learn and optimize the scheduling strategy based on historical scheduling data and current system status; and then building a multi-level scheduling decision framework to dynamically adjust the scheduling strategy and execution plan according to the urgency of the task and the availability of resources; using federated learning algorithms to collaboratively train the prediction model using data from multiple edge nodes; Through the graph neural network algorithm, the topological relationship and communication mode between edge computing nodes are modeled, the data transmission path and scheduling strategy are optimized, and the DAG scheduling algorithm is used according to the dependencies and priorities of edge computing tasks to generate the optimal task execution sequence and resource allocation plan, thereby minimizing the task completion time.

Images