New energy station environment monitoring system and method based on physical sensor

Abstract

The invention relates to the technical field of new energy, and discloses a new energy station environment monitoring system and method based on a physical sensor, and the system comprises a sensor network deployment module which is responsible for deploying a sensor and is used for collecting environment parameters and equipment state parameters of a new energy station; the environmental parameters include temperature, humidity, wind speed, wind direction, illumination intensity, air pressure and rainfall; equipment state parameters include vibration, temperature rise and particulate matter concentration. Through a multi-source data fusion and analysis module, sensor data of different types are integrated by using a weighted fusion algorithm, comprehensive environmental indexes are generated, complex relationships among environmental parameters are mined through multi-dimensional correlation analysis, and compared with a traditional monitoring system which depends on a single data source, the monitoring system has the advantages that the monitoring efficiency is greatly improved. The fusion algorithm significantly improves the reliability of single sensor data, can effectively compensate the error of the sensor, and comprehensively depicts the operation environment of the new energy station.

Description

Technical Field

[0001] The present invention relates to the field of new energy technologies, and in particular to a new energy station environment monitoring system and method based on physical sensors. Background Art

[0002] New energy stations, such as wind farms and photovoltaic power stations, as an important part of clean energy, play an important role in addressing energy shortages and environmental pollution. The optimization of the operating efficiency and equipment performance of new energy stations is highly dependent on environmental conditions. Environmental changes have a significant impact on power generation capacity, equipment stability and operational safety. Therefore, designing an efficient environmental monitoring system to obtain real-time station environment and equipment status information, and to perform intelligent prediction and control, is a core technical requirement to ensure the stable and efficient operation of new energy stations.

[0003] At present, environmental monitoring systems of new energy stations widely use physical sensors for data collection, but existing technologies have the following problems when facing complex environments and dynamic equipment requirements: Existing technologies usually rely on a single sensor or a simple architecture that lacks effective integration between different types of sensors, resulting in the system being unable to accurately capture complex environmental changes. Sensors are also susceptible to environmental interference or their own errors, and a single data source can produce deviations or even errors. In addition, existing systems do not fully consider the correlation between multiple parameters, lack in-depth analysis of the complex interactive relationship between environmental parameters, and are unable to fully characterize the characteristics of the operating environment of new energy stations. Traditional environmental prediction and control strategies are mostly based on rule-based algorithms or simple linear regression models, which have poor adaptability to complex and nonlinear environments. Especially when variables such as wind speed and light intensity fluctuate rapidly, the existing system has insufficient prediction accuracy and cannot generate dynamically adaptive equipment control strategies, resulting in resource waste and even equipment damage, which restricts the development of intelligent management of new energy stations. Most existing environmental monitoring systems are independent monitoring modules, lacking a collaborative working mechanism with the control modules, which enables some systems to generate control parameters and remain at the level of simple equipment adjustment. They are unable to provide real-time feedback and dynamic optimization of equipment execution results. The monitoring and control system that lacks a closed-loop mechanism cannot achieve optimal operation of equipment under complex and changeable environmental conditions, significantly reducing the system's adaptability and long-term operating stability.

[0004] In response to the above problems, the present invention proposes a new energy station environmental monitoring system and method based on physical sensors, aiming to improve the accuracy and reliability of data through multi-source data fusion and intelligent prediction technology. At the same time, a dynamic closed-loop monitoring and control mechanism is constructed to provide comprehensive, intelligent and efficient environmental monitoring and optimization management solutions for new energy stations. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention provides a new energy station environment monitoring system and method based on physical sensors to solve the problems raised in the above background technology.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: a new energy station environment monitoring system based on physical sensors, the system comprising: The sensor network deployment module is responsible for deploying sensors to collect environmental parameters and equipment status parameters of new energy stations; Environmental parameters: temperature, humidity, wind speed, wind direction, light intensity, air pressure and rainfall; Equipment status parameters: vibration, temperature rise and particle concentration; The data acquisition and edge computing module is responsible for the real-time acquisition, preprocessing and preliminary analysis of sensor data. Its specific functions include: Noise reduction: A low-pass filtering algorithm is used to reduce the noise of the collected data. The formula is: , in, For variables The value at time t, For the same variable The value at time t−1, is the smoothing coefficient; Anomaly detection and standardization: Perform range detection and unified dimension processing on data to provide consistent input for the data fusion module. The data standardization formula is: , in, is the standardized data of the ith sensor, is the i-th value after normalization, is the maximum value of all values ​​in the data set. The minimum value of all values ​​in the data set; The data fusion and analysis module performs multi-sensor weighted fusion and multi-dimensional analysis on the collected and pre-processed data. The fused data will be passed to the intelligent prediction and optimization decision module for prediction and analysis. The fusion formula is: , in, is the integrated environmental indicator, is the weight of the i-th sensor, is the standardized data of the i-th sensor, n is the number of sensors; The intelligent prediction and optimization decision module uses the long short-term memory network model to predict environmental changes and generate equipment optimization control parameters. The predicted control parameters are directly input into the automation control and feedback execution module; The monitoring and interaction module is used to provide an interactive interface between users and the system, and to share data with the data acquisition and edge computing modules to provide users with an intuitive interface display; The automated control and feedback execution module dynamically adjusts the operating status of the new energy equipment according to the control parameters generated by the intelligent prediction and optimization decision-making module. Finally, the execution results are fed back to the monitoring and interaction module in real time to form a closed-loop control.

[0007] Preferably, in the intelligent prediction and optimization decision module, the long short-term memory network state update formula is as follows: , , , , , , in, is the hidden state at the current time step, is the input of the current time step, is the memory state of the previous time step, represents the forget gate output at the current time step t, is the weight matrix of the forget gate, is the bias term of the forget gate, represents the activation function, represents the input gate output at the current time step t, is the weight matrix of the input gate, is the bias term of the input gate, is the candidate memory value for the current time step t, is the weight matrix of the candidate memory, is the bias term of the candidate memory, is the hyperbolic tangent activation function, is the output gate value of the current time step, is the weight matrix of the output gate, is the bias term of the output gate, is the memory state of the current time step.

[0008] Preferably, the automatic control and feedback execution module dynamically adjusts the operating state of the new energy equipment according to the control parameters generated by the intelligent prediction and optimization decision module, including: Photovoltaic module angle adjustment, adjust the optimal tilt angle of the photovoltaic module through the formula: , in, is the optimal tilt angle of the photovoltaic module, is the current light intensity, is the maximum light intensity; Fan speed regulation optimizes fan operation based on predicted wind speed data. The algorithm formula is:

[0009] in, is the fan speed, is the current wind speed, For the cut-in wind speed, is the rated wind speed, To cut out the wind speed, is the rated speed, k is the speed adjustment index; like When , the fan does not run and the speed is 0; like When the speed is adjusted by nonlinear function, the wind speed increases and the fan speed increases; like When the wind speed reaches or exceeds the rated wind speed, the fan speed remains constant at the rated speed. ; like The fan stops running.

[0010] Preferably, the arrangement of sensors in the sensor network deployment module meets the following optimization objectives: , in, is the coverage of the ith sensor, and r is the number of sensors.

[0011] Preferably, the data anomaly detection of the data acquisition and edge computing module is based on the triple standard deviation method, and its detection formula is: , in, is the real-time data of the i-th sensor, is the mean of the sampled data, is the standard deviation of the sampled data.

[0012] Preferably, the fusion algorithm in the data fusion and analysis module adopts the Bayesian inference method, and the Bayesian formula is: , in, For a given observation The posterior probability of assuming H is, Given the hypothesis H, The probability of is the prior probability of hypothesis H, Given the hypothesis H, probability.

[0013] Preferably, the fan speed control in the automatic control and feedback execution module is adjusted by a PI controller, and the formula is: , in, is the fan speed, is the speed error, and are the proportional gain and integral gain, represents the time variable used in the integration operation, Represents the cumulative integral of the error.

[0014] Preferably, the speed error of the PI controller is Defined as: , in, is the target speed, is the fan speed.

[0015] Preferably, the visual monitoring function in the monitoring and interaction module includes: Real-time data monitoring, displaying environmental parameters such as temperature, humidity, wind speed, and light intensity through a dynamic dashboard; The sensor layout and status information are displayed on the GIS map.

[0016] A new energy station environment monitoring method based on physical sensors, comprising: Step 1: Deploy a sensor network consisting of multiple types of sensors to collect environmental parameters and equipment status parameters of new energy stations. All sensors are connected to the data acquisition system through a unified interface to provide multi-dimensional raw data input for subsequent steps. Step 2: collect sensor data in real time, and use a low-pass filter algorithm to reduce noise on the raw data to preliminarily eliminate high-frequency noise in the data; Step 3: Perform anomaly detection and standardization on the denoised data, set a range threshold, and remove abnormal data points that exceed the range to provide input for subsequent multi-dimensional data analysis; Step 4: Input the standardized data into the data fusion and analysis module, and fuse the multi-sensor data using the weighted fusion method. After fusion, the comprehensive environmental indicators of the new energy station are obtained, and the indicators are analyzed in multiple dimensions to identify key influencing factors; Step 5: Input the fused data into the intelligent prediction and optimization decision-making module, and predict future environmental changes through the trained long short-term memory network model. The prediction results directly generate the optimized control parameters for equipment operation to adapt to the predicted environmental conditions; Step 6: Input the generated control parameters into the automatic control and feedback execution module. The control module dynamically adjusts the operating state of the new energy equipment according to the input parameters. After the adjustment is completed, the adjusted equipment state is fed back to the monitoring and interaction module to form a closed-loop control process of monitoring-decision-making-control-feedback. Step 7: Receive the sent equipment status information and real-time transmitted environmental monitoring data, and display the data in a visual manner through the user interface. Users can view real-time data, historical data analysis results and equipment operating status in the interface, and can set or adjust relevant monitoring parameters and control strategies to further improve the system's operation and control process.

[0017] The present invention provides a new energy station environment monitoring system and method based on physical sensors, which has the following beneficial effects: 1. The present invention integrates different types of sensor data using a weighted fusion algorithm through a multi-source data fusion and analysis module to generate comprehensive environmental indicators, and mines the complex relationship between various environmental parameters through multi-dimensional correlation analysis. Compared with the traditional monitoring system that relies on a single data source, the fusion algorithm significantly improves the reliability of a single sensor data, and can effectively compensate for the error of the sensor, comprehensively describing the operating environment of the new energy station.

[0018] 2. The present invention applies the long short-term memory network model to the prediction of changes in the new energy station environment and the generation of optimized control parameters. By modeling the time series of historical data, it can accurately predict the future trend of environmental parameter changes and dynamically generate adaptive equipment optimization control strategies. Compared with traditional rule-based or simple linear model control strategies, it achieves deep coupling of prediction and control, greatly improving the system's intelligent decision-making capabilities in complex environments.

[0019] 3. The automated control and feedback execution module of the present invention realizes a dynamic closed-loop control mechanism from data monitoring to equipment operation adjustment. Specifically, the control module dynamically adjusts the wind turbine speed or the angle of the photovoltaic module according to the parameters generated by intelligent prediction. At the same time, the execution result is fed back to the monitoring module in real time, forming a complete closed-loop process of monitoring-analysis-control-feedback. The closed-loop mechanism significantly enhances the system's ability to adapt to environmental changes and can continuously maintain the optimal operating state of the equipment in a dynamic environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 is a system diagram of the present invention; Figure 2 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0021] In order to make the technical personnel in the technical field understand the scheme of the present invention, the technical scheme in the embodiment of the present invention will be clearly and completely described below in combination with the drawings in the embodiment of the present invention. Obviously, the described embodiment is a partial embodiment of the present invention, not a complete embodiment. Based on the embodiment of the present invention, other embodiments obtained by ordinary technicians in the field without creative work should fall within the scope of protection of the present invention.

[0022] The present invention is described in detail below in conjunction with the accompanying drawings: Example: Please see attached Figure 1 and attached Figure 2 The embodiment of the present invention provides a new energy station environment monitoring system based on physical sensors, the system comprising: The sensor network deployment module is responsible for deploying sensors to collect environmental parameters and equipment status parameters of new energy stations: Environmental parameters: temperature, humidity, wind speed, wind direction, light intensity, air pressure and rainfall; Equipment status parameters: vibration, temperature rise and particle concentration; The data acquisition and edge computing module is responsible for the real-time acquisition, preprocessing and preliminary analysis of sensor data. Its specific functions include: Noise reduction: A low-pass filtering algorithm is used to reduce the noise of the collected data. The formula is: , in, For a variable The value at time t, For the same variable The value at time t−1, is the smoothing coefficient; Anomaly detection and standardization: Perform range detection and unified dimension processing on data to provide consistent input for the data fusion module. The data standardization formula is: , in, is the standardized data of the ith sensor, is the i-th value after normalization, is the maximum value of all values ​​in the data set. The minimum value of all values ​​in the data set; The data fusion and analysis module performs multi-sensor weighted fusion and multi-dimensional analysis on the collected and pre-processed data. The fused data will be passed to the intelligent prediction and optimization decision module for prediction and analysis. The fusion formula is: , in, is the integrated environmental indicator, is the weight of the i-th sensor, is the standardized data of the i-th sensor, n is the number of sensors; The intelligent prediction and optimization decision module uses the long short-term memory network model to predict environmental changes and generate equipment optimization control parameters. The predicted control parameters are directly input into the automation control and feedback execution module; The monitoring and interaction module is used to provide an interactive interface between users and the system, and to share data with the data acquisition and edge computing modules to provide users with an intuitive interface display; The automated control and feedback execution module dynamically adjusts the operating status of the new energy equipment according to the control parameters generated by the intelligent prediction and optimization decision-making module. Finally, the execution results are fed back to the monitoring and interaction module in real time to form a closed-loop control.

[0023] The sensor network deployment module optimizes the sensor layout to achieve comprehensive coverage monitoring of environmental parameters and equipment status parameters of new energy stations. At the same time, it monitors equipment vibration, temperature rise and particulate matter concentration D status information, providing a multi-dimensional data source for subsequent data processing; The data acquisition and edge computing module significantly improves the quality and consistency of sensor data through real-time acquisition, noise reduction processing and anomaly detection. The low-pass filtering algorithm is used to reduce the noise of the data, which can effectively filter high-frequency noise and ensure the stability of sensor data. Range detection and standardization processing unify the data of different types of sensors to provide consistent input for subsequent data fusion. In addition, the introduction of edge computing reduces the pressure of data transmission, improves the system processing efficiency, and enables the system to maintain rapid response capabilities in complex environments. The data fusion and analysis module integrates multi-sensor data through a weighted fusion algorithm to generate comprehensive environmental indicators and conducts multi-dimensional analysis on the fused data. The fusion algorithm can compensate for the error of a single sensor and improve the reliability of the data. At the same time, it reveals the complex relationship between environmental parameters through multi-dimensional correlation analysis, providing an important basis for the optimized operation of new energy equipment. Compared with traditional monitoring systems, this module greatly improves the system's ability to analyze complex site environments, laying a solid foundation for accurate prediction and control. The intelligent prediction and optimization decision-making module uses a long-term and short-term memory network model to dynamically predict environmental changes and generate optimized control parameters for equipment operation. By modeling the time series of historical data, the module can accurately predict the future trends of environmental parameters such as wind speed and light intensity, and dynamically generate equipment optimization strategies based on the prediction results to ensure that equipment operation always adapts to complex and changing environmental conditions. Compared with traditional rule-based methods, the intelligent prediction of this module significantly improves the intelligence and efficiency of site management; The monitoring and interaction module displays the system's real-time monitoring data, historical trends and equipment operating status through a visual interface, providing users with an intuitive interactive experience. Users can view key environmental parameters and monitoring alarm information through the interface and adjust the system's monitoring and control strategies. The module's interactive design significantly improves the system's usability and operating efficiency, providing station managers with efficient monitoring tools to ensure that management strategies can quickly respond to the station's actual operating needs; The automated control and feedback execution module implements dynamic closed-loop control of equipment adjustments, and optimizes the operating status of new energy equipment by adjusting the inclination angle of photovoltaic modules and the speed of wind turbines in real time. This module feeds back the equipment adjustment results to the monitoring module in real time, forming a closed-loop process of monitoring-analysis-control-feedback, so that the system can dynamically adapt to environmental changes and continuously maintain the optimal operating status of the equipment, significantly improving the reliability and adaptability of the system, while reducing the need for manual intervention.

[0024] In summary, each module performs its respective functions in data collection, processing, analysis, prediction, control and interaction, and the overall performance of the system is improved through the synergy between modules, effectively solving the shortcomings of traditional new energy station environmental monitoring systems in data accuracy, real-time performance and intelligent control.

[0025] In the intelligent prediction and optimization decision module, the long short-term memory network state update formula is as follows: , , , , , , in, is the hidden state at the current time step, is the input of the current time step, is the memory state of the previous time step, represents the forget gate output at the current time step t, is the weight matrix of the forget gate, is the bias term of the forget gate, represents the activation function, represents the input gate output at the current time step t, is the weight matrix of the input gate, is the bias term of the input gate, is the candidate memory value for the current time step t, is the weight matrix of the candidate memory, is the bias term of the candidate memory, is the hyperbolic tangent activation function, is the output gate value of the current time step, is the weight matrix of the output gate, is the bias term of the output gate, is the memory state of the current time step.

[0026] The state update formula of the long short-term memory network introduces the forget gate, input gate, output gate and other mechanisms, combines the current input, the memory state and hidden state of the previous time step, and dynamically updates the memory state and hidden state of the current time step. Its function is to effectively capture the long-term dependencies in the time series data. At the same time, it forgets irrelevant or redundant information and retains historical information that is useful for the current prediction. This enables the long short-term memory network to model the dynamic changes of environmental parameters in complex time series data, such as the nonlinear change trend of wind speed and light intensity, which is suitable for the prediction task of the new energy station environment. The design of the LSTM network state update formula avoids the problem of gradient vanishing or exploding in traditional recurrent neural networks, making it perform well in long-term dependent time series modeling. The forget gate effectively discards useless historical information, the input gate introduces new information of the current time step, and the output gate regulates the output of the memory state. The LSTM network can generate accurate prediction results under complex and changeable environmental conditions. For new energy stations, the use of LSTM networks can improve the prediction accuracy of environmental parameter change trends and provide efficient and reliable support for equipment operation optimization.

[0027] The automatic control and feedback execution module dynamically adjusts the operating status of new energy equipment according to the control parameters generated by the intelligent prediction and optimization decision module, including: Photovoltaic module angle adjustment, adjust the optimal tilt angle of the photovoltaic module through the formula: , in, is the optimal tilt angle of the photovoltaic module, is the current light intensity, is the maximum light intensity; Fan speed regulation optimizes fan operation based on predicted wind speed data. The algorithm formula is:

[0028] in, is the fan speed, is the current wind speed, For the cut-in wind speed, is the rated wind speed, To cut out the wind speed, is the rated speed, k is the speed adjustment index; like When , the fan does not run and the speed is 0; like When the speed is adjusted by nonlinear function, the wind speed increases and the fan speed increases; like When the wind speed reaches or exceeds the rated wind speed, the fan speed remains constant at the rated speed. ; like The fan stops running.

[0029] Dynamic adjustment of the angle of photovoltaic modules can optimize the utilization of sunlight in different time periods and weather conditions, ensuring the maximum power generation efficiency of the modules. Compared with the fixed angle design, dynamic adjustment can significantly improve the power generation output, while reducing the hot spot effect and long-term loss of photovoltaic modules. In addition, the intelligent adjustment of the system reduces the need for manual intervention and improves the automation and operation efficiency of new energy stations. The wind turbine speed regulation mechanism optimizes wind energy utilization through segmented management of wind speed ranges, stops operation under low wind speed conditions to avoid energy waste, smoothly adjusts the speed within the appropriate wind speed range, improves wind energy conversion efficiency and reduces equipment wear through nonlinear adjustment index k, maintains a constant speed at rated wind speed to ensure power output stability, and automatically shuts down when the wind speed is too high to protect equipment from damage. This regulation mechanism significantly improves the operational safety and energy utilization efficiency of the wind turbine, extends the equipment life, and at the same time ensures the economy and stability of the new energy station.

[0030] The sensor deployment in the sensor network deployment module meets the following optimization goals: , in, is the coverage of the ith sensor, and r is the number of sensors.

[0031] The sensor layout optimization goal in the sensor network deployment module uses the above formula to ensure that the sensor network covers the core area and surrounding areas of the new energy station, and maximizes the spatial coverage of monitoring. By reasonably arranging sensors, the generation of monitoring blind spots is avoided, and key environmental parameters and equipment status can be fully perceived, providing comprehensive and reliable basic data input for subsequent data processing and equipment optimization. By optimizing the layout of sensors, the present invention can reduce redundant deployment and waste of resources, and at the same time, significantly improve the efficiency and accuracy of the monitoring network. Under the premise of meeting full coverage, the optimized layout algorithm can reduce the demand for the number of sensors, reduce system costs and maintenance complexity. In addition, the optimized sensor network can better adapt to complex terrain and environmental changes, ensure that the system can obtain high-quality monitoring data in various scenarios, and improve the reliability and scientificity of the operation and management of new energy sites.

[0032] The data anomaly detection of the data collection and edge computing module is based on the triple standard deviation method, and its detection formula is: , in, is the real-time data of the i-th sensor, is the mean of the sampled data, is the standard deviation of the sampled data.

[0033] The anomaly detection based on the triple standard deviation method in the data acquisition and edge computing module can effectively identify outliers in sensor data through the formula. This method uses the statistical characteristics of the data to mark data points that deviate from the mean by more than three times the standard deviation, and eliminates abnormal data generated by the sensor due to environmental interference or equipment failure, providing reliable input for subsequent data processing and analysis; Anomaly detection through the triple standard deviation method can significantly improve the credibility and accuracy of the data, and avoid the interference of abnormal data on the subsequent analysis and optimization decisions of the system. Compared with the simple threshold detection method, this method can dynamically adapt to the statistical characteristics of the data and has higher robustness and flexibility. In addition, this method is localized in the edge computing module, which effectively reduces the possibility of abnormal data transmission to the central system, thereby reducing the pressure of data transmission and improving the overall real-time and processing efficiency of the system.

[0034] The fusion algorithm in the data fusion and analysis module further adopts the Bayesian inference method. The Bayesian formula is: , in, For a given observation The posterior probability of assuming H is, Given the hypothesis H, The probability of is the prior probability of hypothesis H, Given the hypothesis H, probability.

[0035] The fusion algorithm in the data fusion and analysis module uses the Bayesian inference method and formula to achieve probabilistic inference and fusion of multi-sensor data. This method is based on the conditional probability relationship between observed data and hypotheses, dynamically updates the posterior probability of hypotheses, and extracts the most likely scenario description from multiple environments and device data sources. It can effectively integrate multi-source information and improve the scientificity and accuracy of data fusion. The Bayesian inference method is used for data fusion, which can make full use of the conditional probability relationship of multi-source sensor data to generate reliable environmental indicators and equipment status estimates in the case of redundant or incomplete data. Compared with static fusion algorithms such as weighted average, the Bayesian inference method has dynamic update and adaptive characteristics, and can continuously adjust the fusion results according to real-time data. In addition, this method has good fault tolerance to noise in sensor data, which can significantly improve the robustness and fusion effect of new energy station monitoring systems in complex environments.

[0036] The fan speed control in the automatic control and feedback execution module is adjusted by the PI controller, and its formula is: , in, is the fan speed, is the speed error, and are the proportional gain and integral gain, represents the time variable used in the integration operation, represents the cumulative integral of the error; Speed ​​error of PI controller Defined as: , in, is the target speed, is the fan speed.

[0037] The PI controller has the dual advantages of fast response and high stability in wind turbine speed regulation. It can quickly adjust the speed to match the wind energy utilization demand when the wind speed fluctuates. Through the cumulative error compensation function of the integral term, the system can effectively eliminate long-term deviations and ensure that the wind turbine operates in the target state. Compared with simple proportional control, the PI controller can significantly reduce the overshoot and oscillation of the system, improve the efficiency and stability of the wind turbine operation, extend the life of the equipment, and ensure that the energy output of the station is stable and controllable.

[0038] The visual monitoring functions in the monitoring and interaction module include: Real-time data monitoring, displaying environmental parameters such as temperature, humidity, wind speed, and light intensity through a dynamic dashboard; The sensor layout and status information are displayed on the GIS map.

[0039] A new energy station environment monitoring method based on physical sensors, comprising: Step 1: Deploy a sensor network consisting of multiple types of sensors to collect environmental parameters and equipment status parameters of new energy stations. All sensors are connected to the data acquisition system through a unified interface to provide multi-dimensional raw data input for subsequent steps. Step 2: collect sensor data in real time, and use a low-pass filter algorithm to reduce noise on the raw data to preliminarily eliminate high-frequency noise in the data; Step 3: Perform anomaly detection and standardization on the denoised data, set a range threshold, and remove abnormal data points that exceed the range to provide input for subsequent multi-dimensional data analysis; Step 4: Input the standardized data into the data fusion and analysis module, and fuse the multi-sensor data using the weighted fusion method. After fusion, the comprehensive environmental indicators of the new energy station are obtained, and the indicators are analyzed in multiple dimensions to identify key influencing factors; Step 5: Input the fused data into the intelligent prediction and optimization decision-making module, and predict future environmental changes through the trained long short-term memory network model. The prediction results directly generate the optimized control parameters for equipment operation to adapt to the predicted environmental conditions; Step 6: Input the generated control parameters into the automatic control and feedback execution module. The control module dynamically adjusts the operating state of the new energy equipment according to the input parameters. After the adjustment is completed, the adjusted equipment state is fed back to the monitoring and interaction module to form a closed-loop control process of monitoring-decision-making-control-feedback. Step 7: Receive the sent equipment status information and real-time transmitted environmental monitoring data, and display the data in a visual manner through the user interface. Users can view real-time data, historical data analysis results and equipment operating status in the interface, and can set or adjust relevant monitoring parameters and control strategies to further improve the system's operation and control process.

[0040] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A new energy station environment monitoring system based on physical sensors, characterized in that: The system comprises: The sensor network deployment module is responsible for deploying sensors to collect environmental parameters and equipment status parameters of new energy stations; Environmental parameters: temperature, humidity, wind speed, wind direction, light intensity, air pressure and rainfall; Equipment status parameters: vibration, temperature rise and particle concentration; The data acquisition and edge computing module is responsible for the real-time acquisition, preprocessing and preliminary analysis of sensor data. Its specific functions include: Noise reduction: A low-pass filtering algorithm is used to reduce the noise of the collected data. The formula is: ， in, For variables The value at time t, For the same variable The value at time t−1, is the smoothing coefficient; Anomaly detection and standardization: Perform range detection and unified dimension processing on data to provide consistent input for the data fusion module. The data standardization formula is: , in, is the standardized data of the ith sensor, is the i-th value after normalization, is the maximum value of all values ​​in the data set. The minimum value of all values ​​in the data set; The data fusion and analysis module performs multi-sensor weighted fusion and multi-dimensional analysis on the collected and pre-processed data. The fused data will be passed to the intelligent prediction and optimization decision module for prediction and analysis. The fusion formula is: , in, is the integrated environmental indicator, is the weight of the i-th sensor, is the standardized data of the i-th sensor, n is the number of sensors; The intelligent prediction and optimization decision module uses the long short-term memory network model to predict environmental changes and generate equipment optimization control parameters. The predicted control parameters are directly input into the automation control and feedback execution module; The monitoring and interaction module is used to provide an interactive interface between users and the system, and to share data with the data acquisition and edge computing modules to provide users with an intuitive interface display; The automated control and feedback execution module dynamically adjusts the operating status of the new energy equipment according to the control parameters generated by the intelligent prediction and optimization decision-making module. Finally, the execution results are fed back to the monitoring and interaction module in real time to form a closed-loop control.

2. According to claim 1, a new energy station environment monitoring system based on physical sensors is characterized in that: In the intelligent prediction and optimization decision module, the long short-term memory network state update formula is as follows: ， ， ， ， ， ， in, is the hidden state at the current time step, is the input of the current time step, is the memory state of the previous time step, represents the forget gate output at the current time step t, is the weight matrix of the forget gate, is the bias term of the forget gate, represents the activation function, represents the input gate output at the current time step t, is the weight matrix of the input gate, is the bias term of the input gate, is the candidate memory value for the current time step t, is the weight matrix of the candidate memory, is the bias term of the candidate memory, is the hyperbolic tangent activation function, is the output gate value of the current time step, is the weight matrix of the output gate, is the bias term of the output gate, is the memory state of the current time step.

3. According to claim 1, a new energy station environment monitoring system based on physical sensors is characterized in that: The automatic control and feedback execution module dynamically adjusts the operating state of the new energy equipment according to the control parameters generated by the intelligent prediction and optimization decision module, including: Photovoltaic module angle adjustment, adjust the optimal tilt angle of the photovoltaic module through the formula: ， in, is the optimal tilt angle of the photovoltaic module, is the current light intensity, is the maximum light intensity; Fan speed regulation optimizes fan operation based on predicted wind speed data. The algorithm formula is: in, is the fan speed, is the current wind speed, For the cut-in wind speed, is the rated wind speed, To cut out the wind speed, is the rated speed, k is the speed adjustment index; like When , the fan does not run and the speed is 0; like When the speed is adjusted by nonlinear function, the wind speed increases and the fan speed increases; like When the wind speed reaches or exceeds the rated wind speed, the fan speed remains constant at the rated speed. ; like The fan stops running.

4. According to claim 1, a new energy station environment monitoring system based on physical sensors is characterized in that: The sensor arrangement in the sensor network deployment module meets the following optimization objectives: ， in, is the coverage of the ith sensor, and r is the number of sensors.

5. According to claim 1, a new energy station environment monitoring system based on physical sensors is characterized in that: The data anomaly detection of the data acquisition and edge computing module is based on the triple standard deviation method, and its detection formula is: ， in, is the real-time data of the i-th sensor, is the mean of the sampled data, is the standard deviation of the sampled data.

6. The new energy station environment monitoring system based on physical sensors according to claim 1 is characterized in that: The fusion algorithm in the data fusion and analysis module adopts the Bayesian inference method, and the Bayesian formula is: ， in, For a given observation The posterior probability of assuming H is, Given the hypothesis H, The probability of is the prior probability of hypothesis H, Given the hypothesis H, probability.

7. The new energy station environment monitoring system based on physical sensors according to claim 1 is characterized in that: The fan speed control in the automatic control and feedback execution module is adjusted by a PI controller, and its formula is: ， in, is the fan speed, is the speed error, and are the proportional gain and integral gain, represents the time variable used in the integration operation, Represents the cumulative integral of the error.

8. The new energy station environment monitoring system based on physical sensors according to claim 7 is characterized in that: The speed error of the PI controller Defined as: ， in, is the target speed, is the fan speed.

9. The new energy station environment monitoring system based on physical sensors according to claim 1 is characterized in that: The visual monitoring functions in the monitoring and interaction module include: Real-time data monitoring, displaying environmental parameters such as temperature, humidity, wind speed, and light intensity through a dynamic dashboard; The sensor layout and status information are displayed on the GIS map.

10. A method for monitoring the environment of a new energy station based on a physical sensor, based on a system for monitoring the environment of a new energy station based on a physical sensor according to any one of claims 1 to 9, characterized in that: include: Step 1: Deploy a sensor network consisting of multiple types of sensors to collect environmental parameters and equipment status parameters of new energy stations. All sensors are connected to the data acquisition system through a unified interface to provide multi-dimensional raw data input for subsequent steps. Step 2: collect sensor data in real time, and use a low-pass filter algorithm to reduce noise on the raw data to preliminarily eliminate high-frequency noise in the data; Step 3: Perform anomaly detection and standardization on the denoised data, set a range threshold, and remove abnormal data points that exceed the range to provide input for subsequent multi-dimensional data analysis; Step 4: Input the standardized data into the data fusion and analysis module, and fuse the multi-sensor data using the weighted fusion method. After fusion, the comprehensive environmental indicators of the new energy station are obtained, and the indicators are analyzed in multiple dimensions to identify key influencing factors; Step 5: Input the fused data into the intelligent prediction and optimization decision-making module, and predict future environmental changes through the trained long short-term memory network model. The prediction results directly generate the optimized control parameters for equipment operation to adapt to the predicted environmental conditions; Step 6: Input the generated control parameters into the automatic control and feedback execution module. The control module dynamically adjusts the operating state of the new energy equipment according to the input parameters. After the adjustment is completed, the adjusted equipment state is fed back to the monitoring and interaction module to form a closed-loop control process of monitoring-decision-making-control-feedback. Step 7: Receive the sent equipment status information and real-time transmitted environmental monitoring data, and display the data in a visual manner through the user interface. Users can view real-time data, historical data analysis results and equipment operating status in the interface, and can set or adjust relevant monitoring parameters and control strategies to further improve the system's operation and control process.

Images