Electric energy consumption real-time monitoring management method and system based on Internet of Things

Abstract

The invention discloses an electricity consumption real-time monitoring management method and system based on the Internet of Things, and relates to the technical field of power supply energy consumption monitoring. According to the electricity consumption real-time monitoring and management method based on the Internet of Things, multi-source data of a centralized office place is collected through Internet of Things equipment, edge processing is carried out, a digital twin model is constructed to predict energy consumption, an initial optimization strategy is screened, and the strategy is further optimized in a preset scene in combination with a reinforcement learning algorithm; and finally, the strategy is executed through the Internet of Things platform, correction is performed according to feedback, and meanwhile, the residual life of the equipment is predicted for fault monitoring, so that the core problems of insufficient real-time performance, non-intelligent strategy optimization, equipment fault early warning lag in a centralized office space and the like in the traditional energy consumption management are solved; dynamic optimization of power consumption, accurate monitoring of equipment states and advanced intervention of faults are achieved, and the energy utilization efficiency and the system reliability are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power supply energy consumption monitoring, and specifically to a real-time monitoring and management method and system for electricity energy consumption based on the Internet of Things. Background Art

[0002] There is an urgent need for power safety, energy conservation and emission reduction, and remote management in centralized office places (such as enterprise buildings, schools) and production and manufacturing parks, which requires support for multi-device access (such as air conditioners, large production equipment, lights) and scenario-based modes (such as going to work, lunch break, three-shift operation, etc.). Traditional power distribution systems rely on independent devices (such as electricity meters, leakage switches, etc.), with scattered functions and high costs. Currently, Internet of Things (IoT) and artificial intelligence (AI) technologies have been widely applied in the field of intelligent power consumption. Data collection is achieved through intelligent hardware devices, and load identification, arc analysis, etc. are carried out in combination with AI algorithms to improve power safety and energy efficiency.

[0003] The Chinese invention patent with the publication number CN116385210B discloses a power supply energy consumption monitoring system based on the Internet of Things, including a power supply monitoring device access module, a monitoring period setting module, an electricity consumption monitoring module, an electricity consumption related element collection module, a cloud database, an ordinary electricity consumption trend analysis module, a special electricity consumption trend analysis module, and a power saving demand and power saving potential judgment and identification module. By dividing the whole year into ordinary electricity consumption months and special electricity consumption months, the electricity consumption of the target electricity-consuming enterprise is monitored in the corresponding months, and then the electricity consumption in the ordinary electricity consumption months and special electricity consumption months is classified and analyzed to obtain the electricity consumption trend parameters in the ordinary electricity consumption months and special electricity consumption months, realizing the detailed and targeted monitoring of the enterprise's power supply energy consumption, improving the accuracy of the electricity consumption trend analysis results, being able to provide a true and reliable judgment basis for power saving demand judgment, and effectively reducing the incidence of incorrect judgments.

[0004] However, the existing power supply energy consumption monitoring systems have insufficient real-time performance and non-intelligent strategy optimization. Especially in centralized office places, there is a problem of lagging device fault warning. Summary of the Invention

[0005] Aiming at the deficiencies of the prior art, the present invention provides a real-time monitoring and management method and system for electricity energy consumption based on the Internet of Things, which solves the problems of insufficient real-time performance and non-intelligent strategy optimization existing in traditional energy consumption management. Especially in centralized office places, there is a problem of lagging device fault warning.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: A real-time monitoring and management method for electricity consumption based on the Internet of Things, comprising the following steps: collecting multi-source data of a centralized office location based on Internet of Things devices, and performing edge processing on the multi-source data to obtain processed multi-source data; performing digital twin model modeling based on the processed multi-source data, and performing energy consumption prediction to screen out initial energy consumption optimization strategies; under a preset scenario, screening the initial energy consumption optimization strategies based on a reinforcement learning algorithm to obtain energy consumption optimization strategies; executing the energy consumption optimization strategies through an Internet of Things platform, and correcting the energy consumption optimization strategies based on the feedback results, predicting the remaining life of the device based on a deep learning model, and performing device fault monitoring.

[0007] Further, performing edge processing on the multi-source data to obtain processed multi-source data includes the following steps: obtaining an original data set , including electrical parameters and environmental parameters; performing missing value and outlier processing on the original data set to output a cleaned data set , performing denoising processing on the cleaned data set based on a moving average filtering method to obtain a denoised data set ; performing spatio-temporal alignment on the denoised data set to obtain an aligned data set , calculating derivative indicators based on the aligned data set to obtain derivative indicators of electrical equipment and derivative indicators of environmental status, where the derivative indicators of electrical equipment include load volatility, energy efficiency ratio, and three-phase unbalance degree, and the derivative indicators of environmental status include temperature, humidity, environmental heat capacity, humidity gradient, and arc occurrence frequency; integrating the aligned data set , the derivative indicators of electrical equipment and the derivative indicators of environmental status to obtain a structured energy consumption data set , denoted as processed multi-source data, and storing it in an edge node.

[0008] Further, the edge node analyzes the environment where the electrical equipment is located based on the environment state derivative index, and determines whether the environment where the electrical equipment is located is qualified: performs standardization processing on the environment state derivative index; obtains an environment evaluation value based on the environment evaluation model; if the environment evaluation value is less than the environment evaluation threshold, the environment where the electrical equipment is located is qualified; in the case where the environment where the electrical equipment is located is qualified, analyzes the state of the electrical equipment based on the electrical equipment derivative index and the environment state derivative index of the environment where the electrical equipment is located, and evaluates the operating state of the electrical equipment: compares the environment state derivative index with the environment impact indexes stored in the database one by one to determine the most similar environment impact index, and obtains the allowable deviation value of the load volatility corresponding to the environment impact index stored in the database; processes the electrical equipment derivative index, the reference electrical equipment derivative index, and the allowable deviation value of the load volatility to obtain the electrical equipment state evaluation value, and the reference electrical equipment derivative index includes the reference load volatility, the reference energy efficiency ratio, and the reference three-phase unbalance degree; if the electrical equipment state evaluation value is less than the electrical equipment state evaluation threshold, the operating state of the electrical equipment is qualified.

[0009] Further, the environment evaluation model is expressed as: ; In the formula, is the environment evaluation value, is the standardized temperature, is the standardized humidity, is the standardized environment heat capacity, is the standardized humidity gradient, is the standardized arc occurrence frequency, is the reference standardized temperature, is the reference standardized humidity, is the reference standardized environment heat capacity, is the reference standardized humidity gradient; The method for obtaining the electrical equipment state evaluation value is as follows: ; In the formula, is the load volatility, is the reference load volatility, is the allowable deviation value of the load volatility, is the reference energy efficiency ratio, is the energy efficiency ratio, is the three-phase unbalance degree, is the reference three-phase unbalance degree.

[0010] Further, digital twin model modeling is performed based on the processed multi-source data, including the following steps: obtaining the spatial location information of the electrical equipment and architectural CAD drawings; perform standardization processing on the processed multi-source data and output the standardized multi-source data For the standardized multi-source data perform correlation analysis to obtain features related to energy consumption and obtain features related to energy consumption ; extract the spatial information of electrical equipment from the spatial location information of electrical equipment and architectural CAD drawings ; according to the spatial location information of electrical equipment and architectural CAD drawings, use 3D modeling software to construct a 3D geometric model; map the features related to energy consumption into the 3D geometric model, and establish the spatial topological relationship between electrical equipment according to the spatial information of electrical equipment to obtain an initial 3D digital twin model

[0011] Furthermore, perform energy consumption prediction and screen out the initial energy consumption optimization strategies, including the following steps: use historical features related to energy consumption to train and validate the LSTM model and output the trained LSTM model; obtain the current features related to energy consumption , use the trained LSTM model to predict the load prediction values of electrical equipment for a period of time in the future; simulate the original optimization strategies stored in the database in the initial 3D digital twin model to obtain the load simulation values of electrical equipment for a period of time in the future, obtain the energy consumption change curves corresponding to each original optimization strategy and the response time of electrical equipment; select the original optimization strategies with the energy consumption reduction within the set range and the response time of electrical equipment less than the response threshold, and define them as the initial energy consumption optimization strategies

[0012] Furthermore, integrate the energy consumption status of electrical equipment and the preset scenarios to construct the input data set of the reinforcement learning algorithm , define the state space S, action space A and reward function R of the reinforcement learning algorithm, where , in the formula is the energy consumption reduction is the equipment downtime is 's weight factor is 's weight factor, and the action space A represents the initial energy consumption optimization strategy; use the input data set to train, validate and test the reinforcement learning algorithm and output the trained reinforcement learning algorithm model; input the current energy consumption status of electrical equipment and the preset scenarios into the trained reinforcement learning algorithm model and output the energy consumption optimization strategy

[0013] Furthermore, the energy consumption optimization strategy is executed through the Internet of Things platform, and the energy consumption optimization strategy is corrected based on the feedback results, including the following steps: obtaining the control authority of the power-consuming equipment; parsing the optimization strategy set to extract the specific operation requirements for each power-consuming equipment; verifying whether each operation is allowed to be executed according to the control authority of the power-consuming equipment; generating specific power-consuming equipment control instructions according to the parsed strategy and authority verification results; the Internet of Things platform sends the power-consuming equipment control instructions to the corresponding power-consuming equipment, and the power-consuming equipment feeds back the feedback results to the Internet of Things platform after executing the operation; judging whether the execution effect of the feedback result meets the expectations, if not, dynamically adjusting the optimization strategy set to obtain a corrected optimization strategy set.

[0014] Furthermore, the remaining life of the equipment is predicted based on the deep learning model, and equipment failure monitoring is performed, including the following steps: obtaining equipment operation status data, preprocessing the equipment operation status data, and outputting operation characteristics related to the health status of the equipment ; will run the feature Input into the trained deep learning model and output the remaining life of the equipment; if the remaining life of the equipment is less than the life threshold stored in the database, a fault warning is issued.

[0015] The real-time monitoring and management system of electricity energy consumption based on the Internet of Things is used for the above-mentioned real-time monitoring and management method of electricity energy consumption based on the Internet of Things, including a multi-source data processing module, an optimization strategy initial screening module, an optimization strategy final screening module and an optimization strategy optimization module, wherein: the multi-source data processing module is used to collect multi-source data in centralized office places based on Internet of Things devices, and perform edge processing on the multi-source data to obtain processed multi-source data; the optimization strategy initial screening module is used to model a digital twin model based on the processed multi-source data, and perform energy consumption prediction to screen out an initial energy consumption optimization strategy; the optimization strategy final screening module is used to screen the initial energy consumption optimization strategy based on the reinforcement learning algorithm under a preset scenario to obtain an energy consumption optimization strategy; the optimization strategy optimization module is used to execute the energy consumption optimization strategy through the Internet of Things platform, and correct the energy consumption optimization strategy based on the feedback results, predict the remaining life of the equipment based on the deep learning model, and perform equipment failure monitoring.

[0016] The present invention has the following beneficial effects: The real-time monitoring and management method for electricity consumption based on the Internet of Things. This method collects multi-source data from a centralized office space through Internet of Things devices and performs edge processing. It constructs a digital twin model to predict energy consumption and screen initial optimization strategies, further optimizes the strategies under preset scenarios in combination with the reinforcement learning algorithm, and finally executes the strategies through the Internet of Things platform and corrects them according to the feedback. At the same time, it predicts the remaining life of the equipment for fault monitoring, solves the core problems existing in traditional energy consumption management such as insufficient real-time performance, non-intelligent strategy optimization, especially the lag in equipment fault warning in centralized office spaces, realizes the dynamic optimization of electricity consumption, the accurate monitoring of equipment status, and the early intervention of faults, and significantly improves energy utilization efficiency and system reliability.

[0017] Of course, it is not necessary for any product implementing the present invention to achieve all the above-mentioned advantages simultaneously. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a flowchart of the real-time monitoring and management method for electricity consumption based on the Internet of Things of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] In the embodiments of the present application, through the real-time monitoring and management method and system for electricity consumption based on the Internet of Things, please refer to Figure 1 , the embodiments of the present invention provide a technical solution: The real-time monitoring and management method for electricity consumption based on the Internet of Things includes the following steps: Collect multi-source data based on Internet of Things devices (intelligent air switches, electricity meters, various sensors (such as temperature sensors, humidity sensors, etc.) distributed at different positions in the electricity consumption area), and perform edge processing on the multi-source data to obtain processed multi-source data.

[0020] Obtain the original data set , including electrical parameters and environmental parameters; among them, the electrical parameters are obtained based on intelligent air switches, and data such as voltage and current in the circuit are collected in real time, the electricity meter collects data such as cumulative electricity consumption and real-time power, and the environmental parameters are collected based on various sensors, such as temperature and humidity. These data are recorded in the form of timestamps, reflecting the values of each parameter at different times. Synchronously obtain electrical parameters (voltage, current, power) and environmental parameters (temperature, humidity, arc frequency) to construct a comprehensive equipment operation profile, and solve the problem that traditional single-parameter monitoring cannot reflect the coupled influence of the environment and equipment.

[0021] Process the original data set to handle missing values and outliers, and output the cleaned data set , and perform denoising processing on the cleaned data set based on the moving average filtering method to obtain the denoised data set ; Check whether there are missing values in the data. For a small number of missing values, linear interpolation can be used for filling. Using a statistics-based method, calculate the mean of each parameter μ and standard deviation σ . For data points outside the μ ± kσ range (usually k = 3), they are determined as outliers. For outliers, the historical mean of this parameter can be used for replacement. Use the moving average filtering method to denoise the data.

[0022] Perform spatio-temporal alignment on the denoised data set to obtain the aligned data set . Align multi-source data based on the high-frequency sampling data to solve the time misalignment problem caused by the sampling frequency differences of different devices, and provide a spatio-temporally consistent data source for digital twin modeling.

[0023] Based on the aligned data set , calculate derivative indicators to obtain derivative indicators of electrical equipment and derivative indicators of environmental status. Among them, the derivative indicators of electrical equipment include load volatility, energy efficiency ratio, and three-phase unbalance degree, and the derivative indicators of environmental status include temperature, humidity, environmental heat capacity, humidity gradient, and arc occurrence frequency; Integrate the aligned data set , derivative indicators of electrical equipment and derivative indicators of environmental status to obtain a structured energy consumption data set , denoted as the processed multi-source data and stored in the edge node. Through data governance and feature engineering, a multi-dimensional data system of "equipment - environment - energy consumption" is constructed, which solves the problems of data fragmentation, noise interference, and single feature in traditional energy consumption management, and provides a reliable basis for subsequent intelligent decision-making.

[0024] The edge node analyzes the environment where the electrical equipment is located based on the environment state derivative indicators, and judges whether the environment where the electrical equipment is located is qualified: standardizes the environment state derivative indicators; obtains the environment evaluation value based on the environment evaluation model; if the environment evaluation value is less than the environment evaluation threshold, the environment where the electrical equipment is located is qualified; when the environment where the electrical equipment is located is qualified, analyzes the state of the electrical equipment based on the electrical equipment derivative indicators and the environment state derivative indicators of the environment where the electrical equipment is located, and evaluates the operating state of the electrical equipment: compares the environment state derivative indicators with the environment impact indicators stored in the database one by one (calculated using the Euclidean distance formula), determines the most similar environment impact indicator, and obtains the allowable deviation value of the load volatility corresponding to the environment impact indicator stored in the database; processes the electrical equipment derivative indicators, the specified electrical equipment derivative indicators and the allowable deviation value of the load volatility to obtain the electrical equipment state evaluation value. The specified electrical equipment derivative indicators include the specified load volatility, the specified energy efficiency ratio, and the specified three-phase unbalance degree; if the electrical equipment state evaluation value is less than the electrical equipment state evaluation threshold, the operating state of the electrical equipment is qualified.

[0025] Through the "environment-equipment" two-dimensional dynamic evaluation, the problem of the separation of environmental factors and equipment status in traditional methods is solved, the transformation of fault warning from "post-response" to "early intervention" is realized, and at the same time, a more accurate basis for the health status of equipment is provided for the energy consumption optimization strategy.

[0026] Among them, the environment evaluation model is expressed as: ; In the formula, is the environment evaluation value, is the standardized temperature, is the standardized humidity, is the standardized environmental heat capacity, is the standardized humidity gradient, is the standardized arc occurrence frequency, is the specified standardized temperature, is the specified standardized humidity, is the specified standardized environmental heat capacity, is the specified standardized humidity gradient; The method for obtaining the electrical equipment state evaluation value is as follows: ; In the formula, is the load volatility, is the specified load volatility, is the allowable deviation value of the load volatility, is the specified energy efficiency ratio, is the energy efficiency ratio, is the three-phase unbalance degree, is the specified three-phase unbalance degree.

[0027] Based on the processed multi-source data, a digital twin model is built and energy consumption prediction is carried out to screen out the initial energy consumption optimization strategies.

[0028] Obtain the spatial location information of electrical equipment and the building CAD drawings; perform standardization processing on the processed multi-source data and output the standardized multi-source data , for the standardized multi-source data conduct a correlation analysis to obtain the features related to energy consumption, and obtain the features related to energy consumption , such as power, temperature, time, etc.; combine the equipment spatial location with the building CAD drawings to establish the mapping relationship between the physical space and the digital model. For example, clarify the relative position of the air conditioner and the window through the CAD drawings and analyze the impact of sunlight on the air conditioner energy consumption. Combine the spatial distribution of environmental parameters such as temperature and humidity (such as the temperature and humidity differences between the basement and the top floor) to accurately evaluate the impact of the equipment operating environment on energy consumption. Eliminate the dimension differences of electrical parameters such as voltage and current to ensure the comparability of multi-source data. By calculating the Pearson correlation coefficient, screen out the features strongly related to energy consumption (such as the correlation between the equipment load rate and temperature > 0.8), exclude the interference of redundant data, and improve the model efficiency.

[0029] Extract the spatial information of electrical equipment from the spatial location information of electrical equipment and the building CAD drawings ; according to the spatial location information of electrical equipment and the building CAD drawings, use 3D modeling software (such as Blender, 3dsMax, etc.) to construct a 3D geometric model; use Blender / 3dsMax to construct the 3D model of the equipment to visually display the equipment shape and layout. For example, clearly distinguish the positions of the incoming line cabinet and the outgoing line cabinet in the substation model. Reflect the physical limitations of the equipment installation position through the model (such as poor heat dissipation caused by the ventilation duct obstruction) to provide a basis for energy consumption optimization in the spatial dimension.

[0030] Map the features related to energy consumption to the 3D geometric model, and according to the spatial information of electrical equipment Establish the spatial topological relationships (such as connection relationships, distance relationships, etc.) among electrical equipment to obtain an initial 3D digital twin model, which reflects the relationship between equipment and space, as well as the initial state and attributes of the equipment. Define the electrical connections (such as distribution cabinets connected by bus ducts) and physical distances between equipment to simulate power transmission losses. For example, calculate the resistance losses of long-distance transmission lines. Analyze the impacts such as heat conduction and electromagnetic interference between equipment. For example, the heat accumulation between server cabinets leads to an increase in air-conditioning energy consumption. Through digital twin technology, the deep integration of "physical space - equipment operation - energy consumption data" is achieved, solving the problems of missing spatial dimensions and insufficient analysis of equipment correlation in traditional methods. Its 3D modeling and topological relationship construction provide a more realistic simulation environment for subsequent energy consumption prediction and strategy optimization, significantly improving the system's adaptability to complex scenarios and decision-making accuracy.

[0031] Use historical energy consumption-related features Train and validate the LSTM model, and output the trained LSTM model; use the historical energy consumption-related features Divide them into a training set, a validation set, and a test set, for example, divide them according to a ratio of 7:2:1. The LSTM network consists of an input layer, an LSTM layer, and an output layer. Use the training set to train the selected model, and optimize the model's parameters by minimizing the loss function (such as the mean square error loss function MSE). Use the validation set to validate the trained model, and adjust the hyperparameters of the model (such as the learning rate, the number of neurons in the hidden layer, etc.) according to the validation results to improve the performance of the model. Use the test set to evaluate the final model, and calculate the evaluation metrics (such as the root mean square error RMSE, the mean absolute error MAE, etc.).

[0032] Training the LSTM model with historical energy consumption features can capture the time series patterns of electrical loads (such as weekday / weekend patterns, seasonal fluctuations), integrate multi-dimensional data such as power, temperature, and time, and avoid the one-sidedness of single-index prediction. For example, combine the ambient temperature to predict the air-conditioning load and anticipate the energy consumption peak in advance.

[0033] Obtain the current energy consumption-related features , use the trained LSTM model to predict the load prediction values of electrical equipment for a period of time in the future, for example, predict the load for the next 15 minutes to 1 hour; simulate the original optimization strategies stored in the database in the initial 3D digital twin model to obtain the load simulation values of electrical equipment for a period of time in the future, and obtain the energy consumption change curves and the response times of electrical equipment corresponding to each original optimization strategy, where the energy consumption change curve is expressed as: , where, is the cumulative energy consumption change amount caused by the execution of the original optimization strategy from the starting moment to the t-th time step, The real-time power of the electrical equipment after executing a certain policy at the i-th time step The power of the electrical equipment without executing any policy at the i-th time step, that is, the predicted load value of the electrical equipment by the LSTM model; The original optimization policy with the energy consumption reduction within the set range and the response time of the electrical equipment less than the response threshold is defined as the initial energy consumption optimization policy.

[0034] Preview the original policy in the 3D digital twin model to avoid the risk of actual execution. For example, if the policy of turning off the lighting in a certain area causes insufficient lighting in other areas during the simulation, it can be adjusted in advance. Generate the energy consumption change curve and the equipment response time to accurately measure the policy effect. Select the original optimization policy with the energy consumption reduction within the set range and the response time of the electrical equipment less than the response threshold to balance the energy-saving effect and the equipment feasibility. For example, prefer the frequency converter speed regulation policy (response time 2 minutes) rather than the equipment start-stop policy (response time 8 minutes). Through the combination of "data-driven prediction + virtual scenario verification", the problems in traditional energy consumption management that policy formulation depends on experience and the optimization effect cannot be quantified are solved. The dynamic prediction ability of its LSTM model and the policy preview mechanism of digital twin significantly improve the accuracy and timeliness of energy consumption optimization, laying a foundation for the subsequent in-depth policy optimization of reinforcement learning.

[0035] Under the preset scenario, screen the initial energy consumption optimization policy based on the reinforcement learning algorithm to obtain the energy consumption optimization policy.

[0036] Under the preset scenario, screen the initial energy consumption optimization policy based on the reinforcement learning algorithm to obtain the energy consumption optimization policy, including the following steps: Integrate the energy consumption state of the electrical equipment and the preset scenario to construct the input data set of the reinforcement learning algorithm , define the state space S, action space A and reward function R of the reinforcement learning algorithm, where , in the formula is the energy consumption reduction is the equipment downtime is the weight factor of is the weight factor of The action space A represents the initial energy consumption optimization policy; Use the input data set

[0037] Based on the preset scenarios, create a low-code scenario configuration template. Users can customize new scenarios through simple configuration operations (such as dragging and selecting parameters). Associate the strategies in the preliminary optimization strategy set with different scenarios. For example, in the work mode, select the energy-saving strategies suitable for this scenario. Users use the low-code configuration interface to define new scenarios, such as meeting mode, equipment maintenance mode, etc. The system selects appropriate strategies from the preliminary optimization strategy set according to the user's configuration and generates the corresponding scenario configuration.

[0038] Integrate the energy consumption prediction results with the preset scenarios (such as work, lunch break, three-shift operation, office mode, meeting mode, production mode, etc.) to construct an input data set, ensuring that real-time load and scenario requirements are considered when generating strategies. For example, in the meeting mode, non-essential equipment is preferentially turned off to reduce noise. Define the state space containing the current time, load prediction value, and equipment status, enabling the model to perceive the dynamic changes of the system. For example, adjust the equipment operation strategy in combination with time factors (such as lunch break period). Balance the energy consumption reduction and equipment downtime to avoid sacrificing equipment availability in the pursuit of energy conservation. For example, under the premise of ensuring production continuity, preferentially select short-term energy-saving strategies.

[0039] Execute the energy consumption optimization strategy through the Internet of Things platform, correct the energy consumption optimization strategy based on the feedback results, predict the remaining life of the equipment based on the deep learning model, and conduct equipment fault monitoring.

[0040] Obtain the control permissions of the electrical equipment; analyze the optimization strategy set to extract the specific operation requirements for each electrical equipment; verify whether each operation is allowed to be executed according to the control permissions of the electrical equipment (for operations exceeding the permissions, mark them and generate corresponding warning messages); generate specific control instructions for the electrical equipment according to the analyzed strategies and permission verification results; the Internet of Things platform sends the control instructions for the electrical equipment to the corresponding electrical equipment, and after the electrical equipment executes the operation, it feeds back the results to the Internet of Things platform; judge whether the execution effect of the feedback result meets the expectations, if not, dynamically adjust the optimization strategy set to obtain the corrected optimization strategy set.

[0041] Ensure the legality and security of strategy execution through control permission verification (such as equipment operation white list) to avoid equipment damage caused by misoperations or malicious instructions. Analyze the strategy into control instructions recognizable by the equipment (such as Modbus protocol instructions) to eliminate compatibility problems of equipment from different manufacturers and improve the success rate of instruction execution. Through the "execution - feedback - correction" cycle (response time < 200ms), quickly verify the strategy effect. For example, if the load does not reach the expected value after a certain equipment is turned off, the system automatically adjusts the strategy, shortening the strategy optimization cycle from the traditional hourly level to the minute level.

[0042] Analyze the feedback information of the device to determine whether the execution effect of the strategy meets the expectations. If it is found that there is a deviation between the execution result and the strategy goal, such as the energy consumption reduction not meeting the expectations or the device having abnormal responses, etc., the strategy needs to be corrected. Dynamically adjust the optimization strategy set according to the conclusion of the result analysis. For example, if it is found that turning off some devices causes excessive load on other devices, the start-stop sequence or power distribution of the devices can be adjusted. Based on the feedback data (such as the energy consumption reduction rate, device status), use machine learning algorithms to dynamically update the strategy set to adapt to environmental changes (such as load fluctuations caused by seasonal changes).

[0043] Through the real-time closed-loop of "strategy execution - effect feedback - dynamic correction", the problem of the disconnection between strategy implementation and actual effect in traditional methods is solved; combined with device health management, the transformation from "passive maintenance" to "active prevention" is realized, significantly improving the reliability of the system and the sustainability of energy efficiency optimization.

[0044] Obtain the device operation status data and preprocess the device operation status data to output the operation characteristics related to the device health status ; Input the operation characteristics into the trained deep learning model to output the remaining life of the device; if the remaining life of the device is less than the life threshold stored in the database, a fault warning is issued.

[0045] Collect device operation status data (such as vibration, temperature, current), combine with environmental parameters (such as humidity, heat capacity), and construct a feature set containing more than 20 dimensions, which is more comprehensive than traditional single-parameter monitoring (such as only measuring temperature). Denoise through wavelet transform and screen key features (such as bearing temperature gradient, time-varying coefficient of winding resistance) by mutual information method, and use LSTM or Transformer network to learn the long-term trend of device performance degradation. Input electrical parameters, mechanical vibration signals and environmental data at the same time to solve the problem that traditional methods cannot associate multi-dimensional factors. For example, combine humidity gradient to predict the probability of insulation failure.

[0046] The real-time monitoring and management system of electricity energy consumption based on the Internet of Things is used for the above-mentioned real-time monitoring and management method of electricity energy consumption based on the Internet of Things, including a multi-source data processing module, an optimization strategy initial screening module, an optimization strategy final screening module and an optimization strategy optimization module, wherein the multi-source data processing module is used to collect multi-source data in centralized office places based on Internet of Things devices, and perform edge processing on the multi-source data to obtain processed multi-source data; the optimization strategy initial screening module is used to model a digital twin model based on the processed multi-source data, and perform energy consumption prediction to screen out an initial energy consumption optimization strategy; the optimization strategy final screening module is used to screen the initial energy consumption optimization strategy based on the reinforcement learning algorithm under a preset scenario to obtain an energy consumption optimization strategy; the optimization strategy optimization module is used to execute the energy consumption optimization strategy through the Internet of Things platform, and to correct the energy consumption optimization strategy based on the feedback results, predict the remaining life of the equipment based on the deep learning model, and perform equipment failure monitoring.

[0047] An electronic device comprises: a processor; and a memory, wherein computer program instructions are stored in the memory, and when the computer program instructions are executed by the processor, the processor executes the real-time monitoring and management method for electricity consumption based on the Internet of Things as described above.

[0048] A computer-readable storage medium is used to store a program, and when the program is executed by a processor, it implements the real-time monitoring and management method of electricity consumption based on the Internet of Things as described above.

[0049] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0050] The present invention is described with reference to flowcharts and / or block diagrams of systems, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0051] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means that implement the function specified in one or more of the blocks and / or steps of the flowchart. Figure 1 in one or more of the blocks Figure 1 and / or steps of the flowchart.

[0052] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable apparatus provide steps for implementing the function specified in one or more of the blocks and / or steps of the flowchart. Figure 1 in one or more of the blocks Figure 1 and / or steps of the flowchart.

[0053] Although the preferred embodiments of the present invention have been described, additional changes and modifications to these embodiments can be made by those skilled in the art once they learn of the basic inventive concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0054] It is obvious that those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims

1. A real-time monitoring and management method for electricity consumption based on the Internet of Things, characterized in that, The following steps are involved: Collect multi-source data from centralized offices based on IoT devices, and perform edge processing on the multi-source data to obtain processed multi-source data; Based on the processed multi-source data, digital twin model is built, energy consumption is predicted, and the initial energy consumption optimization strategy is selected; In the preset scenario, the initial energy consumption optimization strategy is screened based on the reinforcement learning algorithm to obtain the energy consumption optimization strategy; Energy consumption optimization strategies are executed through the Internet of Things platform and modified based on feedback results. The remaining life of equipment is predicted based on deep learning models, and equipment failure monitoring is performed.

2. The real-time monitoring and management method of power consumption based on the Internet of Things according to claim 1, characterized in that Performing edge processing on multi-source data to obtain processed multi-source data includes the following steps: Obtain the original data set , including electrical parameters and environmental parameters; For the original data set perform missing value and outlier processing, and output the cleaned data set , based on the moving average filtering method, perform denoising processing on the cleaned data set to obtain the denoised data set ; For the denoised data set perform spatio-temporal alignment to obtain the aligned data set , and based on the aligned data set calculate derived indicators to obtain derived indicators of electrical equipment and derived indicators of environmental status. The derived indicators of electrical equipment include load volatility, energy efficiency ratio, and three-phase unbalance degree, and the derived indicators of environmental status include temperature, humidity, environmental heat capacity, humidity gradient, and arc occurrence frequency; The aligned data set is integrated with the derived indicators of the electrical equipment and the derived indicators of the environmental status to obtain a structured energy consumption data set , which is denoted as the processed multi-source data and stored in the edge node.

3. The real-time monitoring and management method of electricity consumption based on the Internet of Things according to claim 2, characterized in that The edge node analyzes the environment of the power-consuming equipment based on the environmental status derived indicators to determine whether the environment of the power-consuming equipment is qualified: Standardize the derived indicators of environmental status; Based on the environmental assessment model, the environmental assessment value is obtained; If the environmental assessment value is less than the environmental assessment threshold, the environment in which the electrical equipment is located is qualified; When the environment in which the electrical equipment is located is qualified, the state of the electrical equipment is analyzed based on the derived indicators of the electrical equipment and the derived indicators of the environmental state of the environment in which the electrical equipment is located, and the operating state of the electrical equipment is evaluated: Compare the environmental status derived indicators with the environmental impact indicators stored in the database one by one, determine the most similar environmental impact indicator, and obtain the load fluctuation rate allowable deviation value stored in the database corresponding to the environmental impact indicator; The derived index of the power-consuming equipment, the parameter derived index of the power-consuming equipment and the allowable deviation value of the load fluctuation rate are processed to obtain the state evaluation value of the power-consuming equipment, and the parameter derived index of the power-consuming equipment includes the parameter load fluctuation rate, the parameter energy efficiency ratio and the parameter three-phase imbalance degree; If the electrical equipment status assessment value is less than the electrical equipment status assessment threshold, the operating status of the electrical equipment is qualified.

4. The real-time monitoring and management method of electricity consumption based on the Internet of Things according to claim 3, characterized in that The environmental assessment model is expressed as: ; In the formula, is the environmental assessment value, is the standardized temperature, is the standardized humidity, is the standardized environmental heat capacity, is the standardized humidity gradient, is the standardized arc occurrence frequency, is the calibrated standardized temperature, is the calibrated standardized humidity, is the calibrated standardized environmental heat capacity, is the calibrated standardized humidity gradient; Electrical equipment status evaluation value The acquisition method is as follows: ； Wherein, is the load volatility, is the specified load volatility, is the allowable deviation value of the load volatility, is the specified energy efficiency ratio, is the energy efficiency ratio, is the three-phase unbalance degree, is the specified three-phase unbalance degree.

5. The real-time monitoring and management method of power consumption based on the Internet of Things according to claim 1, characterized in that The digital twin model is built based on the processed multi-source data, including the following steps: Obtain the spatial location information of the electrical equipment and the building CAD drawings; Standardize the processed multi-source data and output the standardized multi-source data , for the standardized multi-source data Perform a correlation analysis to obtain features related to energy consumption, and obtain features related to energy consumption ; From the spatial location information of the electrical equipment and extract the spatial information of the electrical equipment from the building CAD drawings ; According to the spatial location information of the electrical equipment and the building CAD drawings, use 3D modeling software to construct a 3D geometric model; Map the characteristics related to energy consumption into the 3D geometric model, and establish the spatial topological relationship between electrical equipment according to the spatial information of electrical equipment, so as to obtain the initialized 3D digital twin model.

6. The real-time monitoring and management method of electricity consumption based on the Internet of Things according to claim 5, characterized in that, Carry out energy consumption forecasting and select the initial energy consumption optimization strategy, including the following steps: Energy consumption-related features of usage history Train and validate the LSTM model, and output the trained LSTM model; Obtain the current energy consumption related features , and use the trained LSTM model to predict the load prediction value of electrical equipment for a period of time in the future; The original optimization strategies stored in the database are simulated in the initialized three-dimensional digital twin model to obtain the load simulation value of the power equipment in the future period, and the energy consumption change curve and power equipment response time corresponding to each original optimization strategy are obtained; The original optimization strategy whose energy consumption reduction is within the set range and whose power consumption equipment response time is less than the response threshold is selected and defined as the initial energy consumption optimization strategy.

7. The real-time monitoring and management method of power consumption based on the Internet of Things according to claim 1, characterized in that In the preset scenario, the initial energy consumption optimization strategy is screened based on the reinforcement learning algorithm to obtain the energy consumption optimization strategy, including the following steps: Integrate the energy consumption status of the electrical equipment and the preset scenarios to construct the input data set for the reinforcement learning algorithm , define the state space S, action space A, and reward function R of the reinforcement learning algorithm, where , in the formula is the energy consumption reduction amount, is the equipment downtime, is 's weight factor, is 's weight factor, and the action space A represents the initial energy consumption optimization strategy; Use the input data set Train, validate, and test the reinforcement learning algorithm, and output the trained reinforcement learning algorithm model; The current energy consumption status of electrical equipment and preset scenarios are input into the trained reinforcement learning algorithm model to output energy consumption optimization strategies.

8. The real-time monitoring and management method of electricity consumption based on the Internet of Things according to claim 1, characterized in that The energy consumption optimization strategy is implemented through the IoT platform, and the energy consumption optimization strategy is modified based on the feedback results, including the following steps: Obtain control authority over electrical equipment; Parse the optimization strategy set to extract specific operation requirements for each electrical device; Verify whether each operation is allowed to be executed according to the control authority of the power-consuming equipment; Generate specific power equipment control instructions based on the parsed strategy and permission verification results; The Internet of Things platform sends the control instructions of the power-consuming equipment to the corresponding power-consuming equipment. After the power-consuming equipment executes the operation, it feeds back the result to the Internet of Things platform; Judge whether the execution effect of the feedback result meets the expectation. If it does not meet the expectation, dynamically adjust the optimization strategy set to obtain the corrected optimization strategy set.

9. The real-time monitoring and management method of electricity consumption based on the Internet of Things according to claim 1, characterized in that Predict the remaining life of the equipment based on the deep learning model and conduct equipment fault monitoring, including the following steps: Obtain the device operation status data, preprocess the device operation status data, and output the operation characteristics related to the device health status ; Input the operating characteristics into the trained deep learning model to output the remaining life of the output device; If the remaining life of the equipment is less than the life threshold stored in the database, a fault warning is issued.

10. An Internet of Things-based real-time monitoring and management system for electricity consumption, which is used for the Internet of Things-based real-time monitoring and management method according to any one of claims 1-9, characterized in that, It includes a multi-source data processing module, an initial screening module for optimization strategies, a final screening module for optimization strategies, and an optimization module for optimization strategies, where: The multi-source data processing module is used to collect multi-source data of the centralized office site based on the Internet of Things devices and perform edge processing on the multi-source data to obtain processed multi-source data; The initial screening module for optimization strategies is used to build a digital twin model based on the processed multi-source data, conduct energy consumption prediction, and screen out the initial energy consumption optimization strategies; The final screening module for optimization strategies is used to screen the initial energy consumption optimization strategies based on the reinforcement learning algorithm under a preset scenario to obtain the energy consumption optimization strategies; The optimization module for optimization strategies is used to execute the energy consumption optimization strategies through the Internet of Things platform, correct the energy consumption optimization strategies based on the feedback results, predict the remaining life of the equipment based on the deep learning model, and conduct equipment fault monitoring.

Images