Energy management method and device, storage medium and electronic equipment

Abstract

The invention relates to an energy management method, an energy management device, a storage medium and electronic equipment. The energy management method comprises the steps that environment data, personnel activity data and equipment operation data are collected in real time through a multi-dimensional sensor network, and the sensor network comprises an illumination sensor, a temperature sensor, a millimeter wave radar personnel activity sensor and an equipment operation sensor; the collected environment data, personnel activity data and equipment operation data are transmitted to a data analysis module through a wired and wireless mixed Internet of Things transmission network; the mixed neural network model based on deep learning analyzes the environment data, the personnel activity data and the equipment operation data to generate an optimal power distribution scheme; and the control instruction is generated according to the optimal power distribution scheme, and the control instruction is issued to the electric equipment through the Internet of Things transmission network, so that more comprehensive and real-time mastering of the energy use condition is facilitated, the energy use is scientifically and dynamically adjusted, and further the continuity of power supply is guaranteed.

Description

Technical Field

[0001] This disclosure relates to the field of energy management technology, and in particular to an energy management method, energy management device, storage medium and electronic device. Background Technology

[0002] In today's society, the efficient use and rational management of energy have become crucial aspects of sustainable development. Traditional energy management methods rely heavily on manual experience and simple timed controls, which have many limitations. In terms of energy consumption monitoring, data collection methods are limited in scope and accuracy, making it difficult to comprehensively and in real-time grasp the energy usage situation. Furthermore, the formulation of energy optimization strategies lacks scientific basis and dynamic adjustment capabilities, affecting power supply stability and energy utilization efficiency. Moreover, the response is slow in the event of faults, making it difficult to ensure the continuity of power supply. Summary of the Invention

[0003] In view of this, the present disclosure aims to provide an energy management method, an energy management device, a storage medium, and an electronic device.

[0004] The technical solution disclosed herein is implemented as follows:

[0005] Firstly, this disclosure provides an energy management method.

[0006] The energy management method provided in this disclosure includes:

[0007] The system collects environmental data, personnel activity data, and equipment operation data in real time through a multi-dimensional sensor network, which includes light sensors, temperature sensors, millimeter-wave radar personnel activity sensors, and equipment operation sensors.

[0008] The collected environmental data, personnel activity data, and equipment operation data are transmitted to the data analysis module via a wired and wireless hybrid Internet of Things (IoT) transmission network.

[0009] A hybrid neural network model based on deep learning analyzes the environmental data, personnel activity data, and equipment operation data to generate the optimal power distribution scheme.

[0010] Control commands are generated based on the optimal power distribution scheme, and then transmitted to electrical equipment via the Internet of Things (IoT) transmission network to achieve energy management.

[0011] In some embodiments, the deep learning-based hybrid neural network model analyzes the environmental data, personnel activity data, and equipment operation data to generate an energy optimization strategy, including:

[0012] A hybrid neural network model based on deep learning is used to analyze the environmental data, personnel activity data, and equipment operation data to determine the patterns of personnel activity, ambient light intensity, indoor and outdoor temperatures, and equipment operating status.

[0013] Based on the aforementioned personnel activity patterns, ambient light intensity, indoor and outdoor temperatures, and equipment operating status, the equipment operating mode and equipment switching time are dynamically adjusted; wherein, the equipment includes at least lighting equipment and air conditioning systems.

[0014] In some embodiments, the hybrid neural network model includes a hybrid load prediction model;

[0015] The real-time acquisition of device operation data through a multi-dimensional sensor network includes:

[0016] The operating parameters of the power distribution system are monitored in real time through smart meters and load sensors;

[0017] The deep learning-based hybrid neural network model analyzes the environmental data, personnel activity data, and equipment operation data to generate energy optimization strategies, including:

[0018] The equipment operation data is analyzed based on the hybrid load forecasting model to predict the future load change trend of the power distribution system and obtain the forecast results; wherein, the hybrid load forecasting model includes a combination model of time series model ARIMA and support vector regression SVR.

[0019] Based on the prediction results and the power distribution system topology, combined with the environmental data and the personnel activity data, the optimal power distribution scheme is generated through a genetic algorithm.

[0020] In some embodiments, the step of generating the optimal power distribution scheme using a genetic algorithm based on the prediction results and the power distribution system topology, combined with the environmental data and the personnel activity data, includes:

[0021] The power distribution scheme is encoded as a chromosome; wherein the power distribution scheme includes branch power allocation ratio, transformer tap position and reactive power compensation status gene position;

[0022] The population size, crossover probability, and mutation probability are set, and the optimal solution of the power distribution scheme is found through iterative evolution; among them, the branch current, transformer load rate, and power factor range are constrained during the power distribution scheme optimization process.

[0023] In some embodiments, the real-time acquisition of environmental data through a multi-dimensional sensor network includes:

[0024] The ambient light intensity is collected based on the light sensor; wherein the light sensor is installed in a grid layout.

[0025] Human activity data is collected based on the aforementioned human activity sensor; wherein, the human activity sensor uses a combination of millimeter-wave radar and Bluetooth Low Energy positioning technology to collect human activity data.

[0026] In some embodiments, the power distribution scheme satisfies the following constraints:

[0027] The current in the distribution branch circuit shall not exceed the rated current;

[0028] The transformer load rate does not exceed the predetermined value;

[0029] By controlling the activation and deactivation of reactive power compensation equipment, the power factor of the power system can be maintained within a predetermined threshold range.

[0030] In some embodiments, the training steps of the hybrid neural network model include:

[0031] The historical data is cleaned and normalized, and then divided into a training set and a validation set.

[0032] The hybrid neural network model is trained based on the training set, and the model parameters of the hybrid neural network model are adjusted using the backpropagation algorithm and the stochastic gradient descent optimization algorithm.

[0033] The model accuracy of the hybrid neural network model is evaluated using the validation set to optimize the convolution kernel size and the number of LSTM memory units.

[0034] Secondly, this disclosure provides an energy management device, comprising:

[0035] The data acquisition module is used to collect environmental data, personnel activity data and equipment operation data in real time through a multi-dimensional sensor network, which includes a light sensor, a temperature sensor, a millimeter-wave radar personnel activity sensor and an equipment operation sensor.

[0036] The data transmission module is used to transmit the collected environmental data, personnel activity data, and equipment operation data to the data analysis module through a wired and wireless hybrid Internet of Things (IoT) transmission network.

[0037] The data analysis module is used to analyze the environmental data, personnel activity data, and equipment operation data based on a hybrid neural network model of deep learning, and generate the optimal power distribution scheme.

[0038] The power distribution scheme generation module is used to generate control commands based on the optimal power distribution scheme, and send the control commands to electrical equipment through the Internet of Things transmission network to realize energy management.

[0039] Thirdly, this disclosure provides a computer-readable storage medium having an energy management program stored thereon, which, when executed by a processor, implements the energy management method described in the first aspect above.

[0040] Fourthly, this disclosure provides an electronic device, including a memory, a processor, and an energy management program stored in the memory and executable on the processor, wherein when the processor executes the energy management program, it implements the energy management method described in the first aspect above.

[0041] The energy management method according to embodiments of this disclosure includes: real-time collection of environmental data, personnel activity data, and equipment operation data through a multi-dimensional sensor network, the sensor network including a light sensor, a temperature sensor, a millimeter-wave radar personnel activity sensor, and an equipment operation sensor; transmission of the collected environmental data, personnel activity data, and equipment operation data to a data analysis module via a wired and wireless hybrid Internet of Things (IoT) transmission network; analysis of the environmental data, personnel activity data, and equipment operation data based on a deep learning hybrid neural network model to generate an optimal power distribution scheme; generation of control commands based on the optimal power distribution scheme, and distribution of the control commands to electrical equipment via the IoT transmission network to achieve energy management. In this application, during energy management, various data are collected through multiple sensors, including environmental data, personnel activity data, and equipment operation data. During energy allocation in energy management, the environmental data, personnel activity data, and equipment operation data are analyzed using a deep learning hybrid neural network model to generate an optimal power distribution scheme. The power distribution scheme is determined through comprehensive analysis of the environmental data, personnel activity data, and equipment operation data, which facilitates a more comprehensive and real-time understanding of energy usage, allows for scientific and dynamic adjustment of energy use, and ultimately helps ensure the continuity of power supply.

[0042] Additional aspects and advantages of this disclosure will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this disclosure. Attached Figure Description

[0043] Figure 1 This is a flowchart illustrating an energy management method according to an exemplary embodiment;

[0044] Figure 2 This is a schematic diagram of an energy management device structure according to an exemplary embodiment. Detailed Implementation

[0045] Embodiments of this disclosure are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this disclosure, and should not be construed as limiting this disclosure.

[0046] In today's society, the efficient use and rational management of energy have become crucial aspects of sustainable development. Traditional energy management methods rely heavily on manual experience and simple timed controls, which have many limitations. In terms of energy consumption monitoring, data collection methods are limited in scope and accuracy, making it difficult to comprehensively and in real-time grasp the energy usage situation. Furthermore, the formulation of energy optimization strategies lacks scientific basis and dynamic adjustment capabilities, affecting power supply stability and energy utilization efficiency. Moreover, the response is slow in the event of faults, making it difficult to ensure the continuity of power supply.

[0047] In view of the above situation, this disclosure provides an energy management method. Figure 1 This is a flowchart illustrating an energy management method according to an exemplary embodiment. Figure 1 As shown, this energy management method includes:

[0048] Step 10: Collect environmental data, personnel activity data, and equipment operation data in real time through a multi-dimensional sensor network, wherein the sensor network includes a light sensor, a temperature sensor, a millimeter-wave radar personnel activity sensor, and an equipment operation sensor;

[0049] Step 11: Transmit the collected environmental data, personnel activity data, and equipment operation data to the data analysis module through a wired and wireless hybrid Internet of Things (IoT) transmission network;

[0050] Step 12: Analyze the environmental data, personnel activity data, and equipment operation data based on a deep learning-based hybrid neural network model to generate the optimal power distribution scheme;

[0051] Step 13: Generate control commands based on the optimal power distribution scheme, and send the control commands to the electrical equipment through the Internet of Things transmission network to realize energy management.

[0052] In this exemplary embodiment, ambient light intensity can be collected by a light sensor, indoor and outdoor building temperatures can be collected by a temperature sensor, personnel activity data can be collected by a millimeter-wave radar personnel activity sensor, and equipment operation status data can be collected by an equipment operation sensor, etc. The collected environmental data, personnel activity data, and equipment operation data are then transmitted to a data analysis module via a wired and wireless hybrid Internet of Things (IoT) transmission network.

[0053] A hybrid neural network model based on deep learning analyzes the environmental data, personnel activity data, and equipment operation data to generate the optimal power distribution scheme.

[0054] Finally, control commands are generated based on the optimal power distribution scheme, and these commands are sent to electrical equipment via the Internet of Things (IoT) transmission network to achieve energy management.

[0055] In this application, energy management involves collecting various data from multiple sensors, including environmental data, personnel activity data, and equipment operation data. During energy allocation within the energy management process, a hybrid neural network model based on deep learning analyzes these data to generate an optimal power distribution plan. This plan, determined through comprehensive analysis of environmental, personnel activity, and equipment operation data, facilitates a more comprehensive and real-time understanding of energy usage, enabling scientific and dynamic adjustments to energy consumption and ultimately ensuring the continuity of power supply.

[0056] In some embodiments, the deep learning-based hybrid neural network model analyzes the environmental data, personnel activity data, and equipment operation data to generate an energy optimization strategy, including:

[0057] A hybrid neural network model based on deep learning is used to analyze the environmental data, personnel activity data, and equipment operation data to determine the patterns of personnel activity, ambient light intensity, indoor and outdoor temperatures, and equipment operating status.

[0058] Based on the aforementioned personnel activity patterns, ambient light intensity, indoor and outdoor temperatures, and equipment operating status, the equipment operating mode and equipment switching time are dynamically adjusted; wherein, the equipment includes at least lighting equipment and air conditioning systems.

[0059] In this exemplary embodiment, the ambient light intensity and indoor and outdoor temperature can be determined by analyzing the environmental data through a hybrid neural network model; the activity patterns of personnel can be determined by analyzing the activity data through a hybrid neural network model; and the operating status of equipment can be determined by analyzing the equipment operation data through a hybrid neural network model.

[0060] In this exemplary embodiment, the operating mode and on / off time of the equipment are dynamically adjusted based on the patterns of personnel activity, ambient light intensity, indoor and outdoor temperatures, and equipment operating status. The equipment includes at least lighting equipment and an air conditioning system. For example, based on the patterns of personnel activity, ambient light intensity, indoor and outdoor temperatures, and equipment operating status, a more suitable lighting mode, such as warm-colored or cool-colored lighting, can be dynamically adjusted for the lighting equipment. Alternatively, based on the patterns of personnel activity, ambient light intensity, indoor and outdoor temperatures, and equipment operating status, the operating mode of the air conditioner can be dynamically adjusted. For example, when personnel activity is dense and frequent, the air conditioner can be adjusted to a cooling mode with a larger cooling airflow; when personnel activity is sparse, the air conditioner can be adjusted to a cooling mode with a smaller cooling airflow. For example, the air conditioner can be set to turn off when the indoor and outdoor temperatures are low and to turn on when the indoor and outdoor temperatures are high. This allows for more rational energy management.

[0061] In some embodiments, the hybrid neural network model includes a hybrid load prediction model;

[0062] The real-time acquisition of device operation data through a multi-dimensional sensor network includes:

[0063] The operating parameters of the power distribution system are monitored in real time through smart meters and load sensors;

[0064] The deep learning-based hybrid neural network model analyzes the environmental data, personnel activity data, and equipment operation data to generate energy optimization strategies, including:

[0065] The equipment operation data is analyzed based on the hybrid load forecasting model to predict the future load change trend of the power distribution system and obtain the forecast results; wherein, the hybrid load forecasting model includes a combination model of time series model ARIMA and support vector regression SVR.

[0066] Based on the prediction results and the power distribution system topology, combined with the environmental data and the personnel activity data, the optimal power distribution scheme is generated through a genetic algorithm.

[0067] In this exemplary embodiment, during the automatic power distribution optimization function, the adjusted operating parameters of the power distribution system (such as voltage, current, power factor, etc.) are fed back to the data analysis center in real time. Based on this feedback information, the center further optimizes and corrects the load forecasting model and power distribution optimization strategy. For example, when there is a significant deviation between the actual load change and the forecast result, the parameters of the load forecasting model or the settings of the intelligent power distribution algorithm are adjusted promptly, enabling the power distribution system to quickly adapt to dynamic load changes and ensuring the stability and efficiency of power supply. This closed-loop feedback control mechanism ensures that the energy management system can continuously optimize and adjust itself according to actual operating conditions, effectively improving the system's adaptability and reliability.

[0068] In some embodiments, the step of generating the optimal power distribution scheme using a genetic algorithm based on the prediction results and the power distribution system topology, combined with the environmental data and the personnel activity data, includes:

[0069] The power distribution scheme is encoded as a chromosome; wherein the power distribution scheme includes branch power allocation ratio, transformer tap position and reactive power compensation status gene position;

[0070] The population size, crossover probability, and mutation probability are set, and the optimal solution of the power distribution scheme is found through iterative evolution; among them, the branch current, transformer load rate, and power factor range are constrained during the power distribution scheme optimization process.

[0071] In this exemplary embodiment, the power distribution scheme is encoded as a chromosome, and the chromosome population is optimized by simulating selection, crossover, and mutation operations in the process of biological evolution to find the optimal power distribution scheme. During the optimization process, for the first time, a variety of complex constraints are comprehensively considered, including current limits for distribution branches, transformer load rate constraints, power factor requirements for reactive power compensation devices, and power supply priorities for different areas.

[0072] Meanwhile, this application can assign corresponding weights to different areas based on their importance (such as important production areas, emergency support areas, etc.), so that the algorithm can prioritize the power supply reliability and stability of important areas when optimizing the power distribution scheme, while minimizing energy loss and maximizing operating efficiency of the power distribution system as a whole.

[0073] This application enables the construction of a complete closed-loop feedback control system for energy management, forming a continuous cyclical optimization process from data acquisition, analysis, strategy formulation, command execution to operation monitoring and feedback. In terms of energy optimization control, the operating status data of the equipment after executing control commands is collected in real time and fed back to the data analysis module. The module dynamically adjusts the energy optimization strategy and related model parameters based on the feedback data. For example, if the lighting equipment, after executing a brightness adjustment command, detects a deviation between the actual light intensity and the expected target through feedback from the light sensor, the system will automatically correct the control command or adjust the model parameters to ensure optimal lighting effect and minimum energy consumption.

[0074] In some embodiments, the real-time acquisition of environmental data through a multi-dimensional sensor network includes:

[0075] The ambient light intensity is collected based on the light sensor; wherein the light sensor is installed in a grid layout.

[0076] Human activity data is collected based on the aforementioned human activity sensor; wherein, the human activity sensor uses a combination of millimeter-wave radar and Bluetooth Low Energy positioning technology to collect human activity data.

[0077] In this exemplary embodiment, a high-precision photodiode array light sensor is used, which features high sensitivity, a wide spectral response range, and low noise. The sensors are installed in a grid layout in each room, corridor, stairwell, and other areas within the building. For example, four light sensors are evenly distributed in a 20-square-meter room, located near the four corners of the room, to ensure accurate sensing of changes in light intensity throughout the area, with a measurement accuracy of ±20 lux.

[0078] In this exemplary embodiment, an advanced millimeter-wave radar people activity sensor is combined with a positioning sensor based on Bluetooth Low Energy (BLE) technology. The millimeter-wave radar sensor can penetrate certain obstacles and accurately detect the movement speed, direction, and stationary state of people. Its detection range can reach a circular area with a radius of 10 meters, and its angular resolution is ±5 degrees. The BLE positioning sensor communicates with the smartphone or smart wearable device carried by the person to achieve accurate positioning of the person, with a positioning accuracy of ±1 meter in an indoor environment. In areas with frequent people activity, such as offices, conference rooms, and shopping malls, a set of people activity sensors is installed every 5 meters to achieve comprehensive monitoring of people's activity patterns.

[0079] In this exemplary embodiment, a platinum resistance temperature sensor (PT100) is selected, which has advantages such as high measurement accuracy (up to ±0.1℃), good stability, and fast response speed. The sensors are installed indoors and outdoors, as well as in key locations such as air conditioning supply and return vents. Outdoor temperature sensors are installed on the roof or unobstructed exterior walls of the building, while indoor temperature sensors are arranged according to a standard of at least one sensor per 30 square meters to ensure accurate acquisition of indoor and outdoor temperature information.

[0080] In this exemplary embodiment, for lighting equipment, current transformers and voltage transformers are installed to monitor the operating current and voltage of the lamps, thereby calculating the power consumption of the lighting equipment; for air conditioning systems, in addition to temperature sensors, pressure sensors and flow sensors are also installed to monitor parameters such as pressure and refrigerant flow of the refrigeration system, as well as information such as fan speed and air volume, to comprehensively understand the operating status of the air conditioning system; for other electrical equipment, such as elevators and production equipment, appropriate power sensors, current sensors, vibration sensors, etc., are installed according to the equipment type. The accuracy of the power sensor can reach ±0.5%, the measurement range of the current sensor is selected according to the rated current of the equipment, and the accuracy is ±1%. The vibration sensor can detect the vibration amplitude and frequency of the equipment to determine the operating stability of the equipment.

[0081] In some embodiments, the power distribution scheme satisfies the following constraints:

[0082] The current in the distribution branch circuit shall not exceed the rated current;

[0083] The transformer load rate does not exceed the predetermined value;

[0084] By controlling the activation and deactivation of reactive power compensation equipment, the power factor of the power system can be maintained within a predetermined threshold range.

[0085] For example, the current in the distribution branch should not exceed the rated current; the transformer load rate should not exceed 80%; and the power factor of the power system should be maintained between 0.9 and 1 by controlling the activation and deactivation of reactive power compensation equipment.

[0086] In this exemplary embodiment, various constraints are fully considered during the optimization process, such as the current of the distribution branch not exceeding its rated current, the transformer load rate not exceeding its safety threshold (e.g., 80%), and the switching of reactive power compensation devices meeting power factor requirements (e.g., the power factor should be maintained between 0.9 and 1). Simultaneously, different weights are assigned according to the importance of different areas (e.g., important production areas, emergency support areas, etc.), prioritizing the reliability and stability of power supply to important areas.

[0087] By assigning appropriate weights to different areas based on their importance (such as key production areas and emergency support areas), the algorithm prioritizes the reliability and stability of power supply to critical areas when optimizing power distribution schemes. Simultaneously, it minimizes energy loss and maximizes operational efficiency across the entire power distribution system. This power distribution optimization method, which deeply integrates intelligent algorithms with multiple constraints, is innovative among similar energy management systems. It effectively addresses the problems of unreasonable power distribution schemes and high operational risks caused by neglecting or incompletely considering constraints in traditional power distribution optimization processes.

[0088] In some embodiments, the training steps of the hybrid neural network model include:

[0089] The historical data is cleaned and normalized, and then divided into a training set and a validation set.

[0090] The hybrid neural network model is trained based on the training set, and the model parameters of the hybrid neural network model are adjusted using the backpropagation algorithm and the stochastic gradient descent optimization algorithm.

[0091] The model accuracy of the hybrid neural network model is evaluated using the validation set to optimize the convolution kernel size and the number of LSTM memory units.

[0092] In this exemplary embodiment, a hybrid forecasting model is employed, combining a time series analysis model (such as the ARIMA model) with a machine learning model (such as the Support Vector Machine Regression model, SVR). The ARIMA model is used to capture the linear time series characteristics of load data, and its parameters (p, d, q) are determined through analysis using the Autocorrelation Function (ACF) and Partial Autocorrelation Function (PACF). For example, for daily load data, analysis determines p = 2, d = 1, and q = 2. The SVR model is used to handle the nonlinear relationships in the load data. It maps the data to a high-dimensional space for regression analysis using a kernel function (such as the Radial Basis Function, RBF). The parameters of the kernel function (such as γ and C) are optimized using cross-validation to improve the model's forecasting accuracy.

[0093] In this exemplary embodiment, a large amount of historical load data can be used to train the constructed hybrid prediction model, with the data divided into a training set and a validation set in a 70%:30% ratio. During training, metrics such as mean squared error (MSE) and mean absolute error (MAE) are used as loss functions, and the model parameters are adjusted using gradient descent or other optimization algorithms to minimize the loss function value. After training and validation, the model achieves an MAE within ±3% when predicting load changes over the next hour, and within ±5% when predicting load changes over the next 24 hours.

[0094] A hybrid neural network model based on deep learning is constructed, integrating the advantages of Convolutional Neural Networks (CNNs) and Long Short-Term Memory Networks (LSTMs). CNNs are primarily used for feature extraction from spatial data, such as analyzing images of light intensity distribution and heat maps of human activity. The size of the convolutional kernels is optimized based on the data characteristics, such as 3x3 or 5x5. Through multiple layers of convolution and pooling operations, key spatial features are extracted from the data. LSTMs, on the other hand, focus on processing time-series data, such as historical energy consumption data, time series of human activity, and temperature change curves. Their internal memory units effectively capture long-term dependencies in the data. The parameters of the forget gate, input gate, and output gate are trained and adjusted using backpropagation and stochastic gradient descent algorithms to ensure that the model can accurately predict future energy demand and human activity trends.

[0095] The model training data sources and preprocessing include: collecting a large amount of historical data, including energy consumption data, environmental data (light intensity, indoor and outdoor temperature, humidity, etc.), personnel activity data (number of people, location, activity trajectory, etc.), and equipment operation data (equipment power, running time, number of start-ups and shutdowns, etc.) for different seasons, date types (weekdays, weekends, holidays), and time periods (daytime, nighttime, peak hours, off-peak hours) under different time periods. This data is preprocessed by first cleaning the data using statistical outlier detection methods, such as the 3σ principle, to remove data points that significantly deviate from the normal range; then, data normalization is performed, mapping different types of data to a unified numerical range, such as [0,1] or [-1,1], to improve the efficiency and accuracy of model training.

[0096] The system includes a data analysis and processing module to analyze and organize the collected initial data. During data cleaning, various methods are employed to handle outliers. For example, light intensity data exceeding three times the normal range is considered anomaly, and mean imputation is used to handle missing data. Data is normalized, mapping different data types to the [0,1] interval for model training. Historical data is used to train and optimize the artificial intelligence algorithm model. For the hybrid neural network model developed for energy optimization strategies, historical energy consumption, environmental data, and personnel activity data are divided into training and validation sets in a 7:3 ratio. During training, the CNN convolutional kernel size is set to alternate between 3x3 and 5x5, and the number of LSTM memory units is determined based on data characteristics. Backpropagation and stochastic gradient descent optimization algorithms are used to adjust model parameters. The model's performance is evaluated and continuously optimized using metrics such as validation set accuracy and mean squared error.

[0097] During system operation, the data acquisition subsystem continuously collects various types of data and transmits them to the data analysis and strategy formulation module. This module formulates energy optimization strategies based on real-time data and trained models. Regarding lighting equipment, when the activity sensor detects sparse crowds and the light sensor indicates sufficient natural light (e.g., during off-hours in an office), the model generates control commands. The lighting controller receives these commands and uses PWM technology to adjust the lamp drive circuits according to the brightness adjustment values ​​specified in the commands, reducing the brightness by 10% every 5 minutes, starting with lamps furthest from people, until some lamps are turned off. When the air conditioning system is running, the data analysis and strategy formulation module uses data from indoor and outdoor temperature sensors and the number of people. For example, if the indoor temperature is above 28℃ and there are few people, it generates a ventilation mode command. The air conditioning intelligent control terminal receives the command, switches the operating mode, and adjusts the fan speed to control the ventilation volume. When there are many people indoors and the outdoor temperature is high, the cooling power increases by 10% for every 10 additional people, and the temperature setpoint is dynamically adjusted between 24-26℃, achieved by controlling the operating status of components such as the compressor and fan. For other electrical equipment, such as production equipment, based on the processing workload and equipment operation sensor data, when the workload decreases, the control module receives instructions to reduce the equipment power output or suspend equipment operation to save energy.

[0098] Meanwhile, the automatic power distribution optimization function continues to operate. The real-time load monitoring subsystem collects data from smart meters and load sensors every second and transmits it to the big data analysis and load forecasting module. This module cleans and extracts features from the data before inputting it into the load forecasting model to predict future load change trends. When a load change is predicted, the power distribution optimization decision and adjustment module uses a genetic algorithm to make power distribution optimization decisions based on the prediction results, combined with the power distribution system topology, capacity limitations, and current operating status. For example, a population size of 50, a crossover probability of 0.7, and a mutation probability of 0.05 are set for 100 generations of evolutionary iterations. During the optimization process, constraints such as distribution branch current not exceeding the rated current, transformer load rate not exceeding 80%, and reactive power compensation device switching meeting the power factor requirement of 0.9-1 are considered, and weights are assigned according to regional importance.

[0099] The generated power distribution scheme is converted into control commands and sent to the power distribution automation equipment. The intelligent switchgear controls the opening and closing of switches to adjust branch power distribution according to the commands. The on-load tap-changing transformer switches taps to adjust the output voltage, and the reactive power compensation controller controls the switching of capacitor banks to regulate reactive power. During the power distribution adjustment process, operating parameters are monitored in real time and fed back to the big data analysis and load forecasting module to optimize and correct the model and strategy, forming a closed-loop control system. This ensures that the power distribution system is always in optimal operating condition, effectively improving energy efficiency, reducing energy consumption and operating costs, and providing stable, efficient, and intelligent energy management services for buildings or industrial sites.

[0100] This disclosure provides an energy management device. Figure 2 This is a schematic diagram of an energy management device structure according to an exemplary embodiment. Figure 2 As shown, the energy management device includes:

[0101] The data acquisition module 20 is used to collect environmental data, personnel activity data and equipment operation data in real time through a multi-dimensional sensor network, which includes a light sensor, a temperature sensor, a millimeter-wave radar personnel activity sensor and an equipment operation sensor.

[0102] The data transmission module 21 is used to transmit the collected environmental data, personnel activity data and equipment operation data to the data analysis module through a wired and wireless hybrid Internet of Things transmission network;

[0103] Data analysis module 22 is used to analyze the environmental data, personnel activity data and equipment operation data based on a deep learning hybrid neural network model to generate the optimal power distribution scheme;

[0104] The power distribution scheme generation module 23 is used to generate control commands based on the optimal power distribution scheme and send the control commands to electrical equipment through the Internet of Things transmission network to realize energy management.

[0105] In this exemplary embodiment, ambient light intensity can be collected by a light sensor, indoor and outdoor building temperatures can be collected by a temperature sensor, personnel activity data can be collected by a millimeter-wave radar personnel activity sensor, and equipment operation status data can be collected by an equipment operation sensor, etc. The collected environmental data, personnel activity data, and equipment operation data are then transmitted to a data analysis module via a wired and wireless hybrid Internet of Things (IoT) transmission network.

[0106] A hybrid neural network model based on deep learning analyzes the environmental data, personnel activity data, and equipment operation data to generate the optimal power distribution scheme.

[0107] Finally, control commands are generated based on the optimal power distribution scheme, and these commands are sent to electrical equipment via the Internet of Things (IoT) transmission network to achieve energy management.

[0108] In this application, energy management involves collecting various data from multiple sensors, including environmental data, personnel activity data, and equipment operation data. During energy allocation within the energy management process, a hybrid neural network model based on deep learning analyzes these data to generate an optimal power distribution plan. This plan, determined through comprehensive analysis of environmental, personnel activity, and equipment operation data, facilitates a more comprehensive and real-time understanding of energy usage, enabling scientific and dynamic adjustments to energy consumption and ultimately ensuring the continuity of power supply.

[0109] This disclosure provides a computer-readable storage medium having an energy management program stored thereon, which, when executed by a processor, implements the energy management methods described in the above embodiments.

[0110] This disclosure provides an electronic device, including a memory, a processor, and an energy management program stored in the memory and executable on the processor. When the processor executes the energy management program, it implements the energy management methods described in the above embodiments.

[0111] It should be noted that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which programs can be printed, because programs can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0112] It should be understood that various parts of this disclosure can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0113] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this disclosure. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0114] In the description of this disclosure, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this disclosure and simplifying the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this disclosure.

[0115] Furthermore, the terms "first," "second," etc., used in the embodiments of this disclosure are for descriptive purposes only and should not be construed as indicating or implying relative importance, or implicitly specifying the number of technical features indicated in this embodiment. Therefore, features defined with terms such as "first" and "second" in the embodiments of this disclosure can explicitly or implicitly indicate that the embodiment includes at least one of those features. In the description of this disclosure, the word "multiple" means at least two or more, such as two, three, four, etc., unless otherwise explicitly specified in the embodiments.

[0116] In this disclosure, unless otherwise explicitly specified or limited in the embodiments, the terms "installation," "connection," "joining," and "fixing," etc., appearing in the embodiments should be interpreted broadly. For example, a connection can be a fixed connection, a detachable connection, or an integral part; it can also be a mechanical connection, an electrical connection, etc. Of course, it can also be a direct connection, or an indirect connection through an intermediate medium, or it can be the internal communication between two components, or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this disclosure based on the specific implementation.

[0117] In this disclosure, unless otherwise expressly specified and limited, "above" or "below" the second feature can mean that the first and second features are in direct contact, or that the first and second features are in indirect contact through an intermediate medium. Furthermore, "above," "on top of," and "over" the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply that the first feature is at a lower horizontal level than the second feature.

[0118] Although embodiments of the present disclosure have been shown and described above, it is to be understood that the above embodiments are exemplary and should not be construed as limiting the present disclosure. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present disclosure.

Claims

1. An energy management method, characterized in that, include: The system collects environmental data, personnel activity data, and equipment operation data in real time through a multi-dimensional sensor network, which includes light sensors, temperature sensors, millimeter-wave radar personnel activity sensors, and equipment operation sensors. The collected environmental data, personnel activity data, and equipment operation data are transmitted to the data analysis module via a wired and wireless hybrid Internet of Things (IoT) transmission network. A hybrid neural network model based on deep learning analyzes the environmental data, personnel activity data, and equipment operation data to generate the optimal power distribution scheme. Control commands are generated based on the optimal power distribution scheme, and then transmitted to electrical equipment via the Internet of Things (IoT) transmission network to achieve energy management.

2. The energy management method according to claim 1, characterized in that, The deep learning-based hybrid neural network model analyzes the environmental data, personnel activity data, and equipment operation data to generate energy optimization strategies, including: A hybrid neural network model based on deep learning is used to analyze the environmental data, personnel activity data, and equipment operation data to determine the patterns of personnel activity, ambient light intensity, indoor and outdoor temperatures, and equipment operating status. Based on the aforementioned personnel activity patterns, ambient light intensity, indoor and outdoor temperatures, and equipment operating status, the equipment operating mode and equipment switching time are dynamically adjusted; wherein, the equipment includes at least lighting equipment and air conditioning systems.

3. The energy management method according to claim 2, characterized in that, The hybrid neural network model includes a hybrid load prediction model; The real-time acquisition of device operation data through a multi-dimensional sensor network includes: The operating parameters of the power distribution system are monitored in real time through smart meters and load sensors; The deep learning-based hybrid neural network model analyzes the environmental data, personnel activity data, and equipment operation data to generate energy optimization strategies, including: The equipment operation data is analyzed based on the hybrid load forecasting model to predict the future load change trend of the power distribution system and obtain the forecast results; wherein, the hybrid load forecasting model includes a combination model of time series model ARIMA and support vector regression SVR. Based on the prediction results and the power distribution system topology, combined with the environmental data and the personnel activity data, the optimal power distribution scheme is generated through a genetic algorithm.

4. The energy management method according to claim 3, characterized in that, Based on the prediction results and the power distribution system topology, combined with the environmental data and the personnel activity data, the optimal power distribution scheme is generated through a genetic algorithm, including: The power distribution scheme is encoded as a chromosome; wherein the power distribution scheme includes branch power allocation ratio, transformer tap position and reactive power compensation status gene position; The population size, crossover probability, and mutation probability are set, and the optimal solution of the power distribution scheme is found through iterative evolution; among them, the branch current, transformer load rate, and power factor range are constrained during the power distribution scheme optimization process.

5. The energy management method according to claim 3, characterized in that, The real-time acquisition of environmental data through a multi-dimensional sensor network includes: The ambient light intensity is collected based on the light sensor; wherein the light sensor is installed in a grid layout. Human activity data is collected based on the aforementioned human activity sensor; wherein, the human activity sensor uses a combination of millimeter-wave radar and Bluetooth Low Energy positioning technology to collect human activity data.

6. The method according to claim 4, characterized in that, The power distribution scheme meets the following constraints: The current in the distribution branch circuit shall not exceed the rated current; The transformer load rate does not exceed the predetermined value; By controlling the activation and deactivation of reactive power compensation equipment, the power factor of the power system can be maintained within a predetermined threshold range.

7. The method according to claim 6, characterized in that, The training steps of the hybrid neural network model include: The historical data is cleaned and normalized, and then divided into a training set and a validation set. The hybrid neural network model is trained based on the training set, and the model parameters of the hybrid neural network model are adjusted using the backpropagation algorithm and the stochastic gradient descent optimization algorithm. The model accuracy of the hybrid neural network model is evaluated using the validation set to optimize the convolution kernel size and the number of LSTM memory units.

8. An energy management device, characterized in that, include: The data acquisition module is used to collect environmental data, personnel activity data and equipment operation data in real time through a multi-dimensional sensor network, which includes a light sensor, a temperature sensor, a millimeter-wave radar personnel activity sensor and an equipment operation sensor. The data transmission module is used to transmit the collected environmental data, personnel activity data, and equipment operation data to the data analysis module through a wired and wireless hybrid Internet of Things (IoT) transmission network. The data analysis module is used to analyze the environmental data, personnel activity data, and equipment operation data based on a hybrid neural network model of deep learning, and generate the optimal power distribution scheme. The power distribution scheme generation module is used to generate control commands based on the optimal power distribution scheme, and send the control commands to electrical equipment through the Internet of Things transmission network to realize energy management.

9. A computer-readable storage medium, characterized in that, It stores an energy management program, which, when executed by a processor, implements the energy management method according to any one of claims 1-7.

10. An electronic device, characterized in that, The device includes a memory, a processor, and an energy management program stored in the memory and executable on the processor. When the processor executes the energy management program, it implements the energy management method according to any one of claims 1-7.

Images