Fireproof early warning method and system for mountain photovoltaic power station
By deploying a multi-type sensor network in mountain photovoltaic power stations, and combining deep learning algorithms with terrain data, the problem of false alarms and missed alarms caused by single monitoring data in existing systems has been solved, enabling accurate assessment and rapid response to fire risks.
Patent Information
- Application Number
- CN202511651814.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-12
- Publication Date
- 2026-01-20
AI Technical Summary
Existing fire early warning systems for mountain photovoltaic power stations rely on a single type of sensor, resulting in limited monitoring data dimensions, a high risk of false alarms and missed alarms, and an inability to accurately assess fire risks, thus failing to meet the fire prevention requirements of mountain photovoltaic power stations.
Multi-sensor networks are used to acquire multi-dimensional monitoring data in real time. Data fusion analysis is performed by combining deep learning algorithms with topographic data and equipment operating status data. Accurate assessment is then conducted using fire risk assessment algorithms, and fire early warning signals are generated.
This improved the comprehensiveness and accuracy of monitoring data, reduced the probability of false alarms and missed alarms, and enabled accurate assessment and rapid response to fire risks.
Smart Images
Figure CN121366465A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of intelligent fire fighting of power stations, in particular to a fire prevention and early warning method and system for mountain photovoltaic power stations. BACKGROUND
[0002] Mountain photovoltaic power stations refer to photovoltaic power stations built in mountainous or hilly areas. Compared with photovoltaic power stations in plain areas, mountain photovoltaic power stations have the characteristics of complex terrain, variable climate and dense vegetation, and once a fire breaks out in the mountains, it will cause the fire to spread rapidly, resulting in greater difficulty in extinguishing.
[0003] At present, the traditional fire prevention and early warning of mountain photovoltaic power stations mostly rely on a single type of sensor, such as monitoring only by temperature sensors or smoke sensors, and the monitoring data dimension is single, which is prone to false positives and false negatives. Some systems that collect multi-sensor monitoring lack accuracy and real-time performance in data fusion and analysis, and are difficult to quickly and accurately assess fire risk. Moreover, the existing early warning system does not take into account the unique topography and equipment operating conditions of the mountains, and cannot comprehensively judge the possibility and danger of fire, resulting in low reliability and effectiveness of the early warning, and it is difficult to meet the actual needs of fire prevention in mountain photovoltaic power stations.
[0004] In view of the above, a fire prevention and early warning method and system for mountain photovoltaic power stations are proposed. SUMMARY
[0005] The purpose of the present application is to provide a fire prevention and early warning method and system for mountain photovoltaic power stations to solve the problems raised in the background.
[0006] To achieve the above-mentioned purpose, the present application provides the following technical scheme: a fire prevention and early warning method for mountain photovoltaic power stations, comprising the following steps:
[0007] Based on the multi-type sensor network deployed in the mountain photovoltaic power station, real-time multi-dimensional monitoring data is obtained, including temperature data, smoke concentration data, image data and environmental wind speed data;
[0008] The obtained multi-dimensional monitoring data is preprocessed to remove noise data and interpolate missing data;
[0009] The preprocessed multi-dimensional monitoring data is input into a preset fusion analysis model, the fusion analysis model is constructed based on a deep learning algorithm, and the data is fused by the following formula:
[0010] wherein F is the fused data value, D is the standard value of the i-th type of monitoring data, is the weight coefficient of the i-th type of monitoring data, b is the bias term, and n is the type of monitoring data.
[0011] Based on the fused data, combined with the topographic and geomorphic data and equipment operation state data of the mountain photovoltaic power station, a preset fire risk assessment algorithm is used for judgment, and the calculation formula of the fire risk assessment algorithm is:
[0012] Wherein, R is the fire risk value, T is the normalized value of temperature data, S is the normalized value of smoke concentration data, I is the fire feature value of image data, V is the influence factor of environmental wind speed data, G is the fire spread coefficient of topographic and geomorphic data, E is the abnormal index of equipment operation state data, is the weight coefficient of corresponding data;
[0013] When it is judged that the fire risk value exceeds the preset risk threshold value, a fire prevention warning signal is generated, and the specific position of the warning area is determined according to the regional division of the mountain photovoltaic power station.
[0014] Preferably, the multi-type sensor network comprises a distributed optical fiber temperature sensor, an infrared smoke sensor, a high-definition monitoring camera and a wind speed and direction sensor, wherein the distributed optical fiber temperature sensor is arranged along the cable of the photovoltaic array for monitoring the cable temperature.
[0015] Preferably, in the pre-processing step of the multi-dimensional monitoring data, wavelet transform denoising algorithm is used for denoising processing of temperature data and smoke concentration data, and the wavelet transform formula is:
[0016] , wherein, is the original signal, is the wavelet coefficient, is the wavelet base function.
[0017] Preferably, the construction process of the fusion analysis model comprises: collecting historical monitoring data and corresponding fire time data, extracting features from the historical monitoring data, learning the features by using a convolutional neural network, and optimizing the weight coefficient and the bias term b by using a back propagation algorithm.
[0018] Preferably, the topographic and geomorphic data comprises altitude, slope, slope direction and vegetation coverage, which is obtained by a geographic information system and used for calculating the fire spread coefficient G, and the calculation formula of the fire spread coefficient is:
[0019] , wherein K is an empirical coefficient, slope is the slope value, is the vegetation coverage.
[0020] Preferably, the equipment operating state data includes the working voltage, working current of the photovoltaic module and the operating parameters of the inverter, which are collected in real time by the SCADA system, and the calculation method of the abnormality index E is:
[0021] wherein, is the real-time value of the jth equipment operating parameter, is the historical average value of the jth equipment operating parameter, is the historical standard deviation of the jth equipment operating parameter, is the importance coefficient of the jth equipment operating parameter, and m is the number of equipment operating parameters.
[0022] Preferably, after the fire prevention warning is generated, a preset emergency handling scheme is called according to the specific position of the warning area and the fire risk value, and the emergency handling scheme includes starting the sprinkler fire extinguishing system, cutting off the power supply of the corresponding area and sending alarm information to the monitoring center.
[0023] A fire prevention warning system for a mountain photovoltaic power station, comprising:
[0024] A data acquisition module for acquiring multi-dimensional monitoring data in real time through a multi-type sensor network;
[0025] A data preprocessing module for preprocessing the acquired multi-dimensional monitoring data;
[0026] A data fusion analysis module for inputting the preprocessed multi-dimensional monitoring data into a fusion analysis model for fusion processing;
[0027] A risk assessment and judgment module for performing fire risk assessment and judgment according to the fused data in combination with topographic and geomorphic data and equipment operating state data;
[0028] A warning generation module for generating a fire prevention warning signal and determining the position of a warning area when the fire risk value exceeds a preset threshold.
[0029] Preferably, the system further comprises an emergency handling module for calling a preset emergency handling scheme according to the position of the warning area and the fire risk value.
[0030] Preferably, the data acquisition module, the data preprocessing module, the data fusion analysis module, the risk assessment and judgment module and the warning generation module are integrated in a central processing unit, and the central processing unit adopts an embedded processor.
[0031] Compared with the prior art, the present application has the following advantages:
[0032] The fire prevention early warning method and system of the mountain photovoltaic power station, through multiple types of sensors, real-time acquisition of temperature, smoke concentration, image, environmental wind speed and other multi-dimensional monitoring data, and aiming at the hot spots of the cable of the mountain photovoltaic power station prone to heating, distributed fiber temperature sensors are arranged along the photovoltaic array cable, precise monitoring of the key parts is realized, the comprehensiveness and accuracy of the monitoring data are improved, and the multi-dimensional monitoring data is preprocessed, the wavelet transform method denoising algorithm and interpolation completion are used to improve the data quality.
[0033] At the same time, a fusion analysis model is constructed based on a deep learning algorithm, historical data features are learned by using a convolutional neural network, parameters are optimized by using a back propagation algorithm, data deep fusion is realized, the fire risk value is calculated by using a fire risk assessment algorithm in combination with topographic and geomorphic data and equipment operation state data, the fire risk can be accurately evaluated, and the false alarm and missed alarm probabilities are reduced. BRIEF DESCRIPTION OF DRAWINGS
[0034] Figure 1 It is a mountain photovoltaic power station fire prevention early warning process schematic diagram of the application;
[0035] Figure 2 It is a data acquisition module process schematic diagram of the application;
[0036] Figure 3 It is a risk assessment and judgment module process schematic diagram of the application;
[0037] Figure 4 It is a data preprocessing module process schematic diagram of the application. DETAILED DESCRIPTION
[0038] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application will be further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application.
[0039] In the description of the present application, it should be understood that the terms "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", "axial", "radial", "circumferential" and the like indicate the orientation or positional relationship shown in the drawings, and are only used to facilitate the description of the present application and simplify the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, therefore it cannot be understood as a limitation of the present application.
[0040] In addition, the terms "first", "second", "third", etc. are used herein only to describe various instances, and do not imply or suggest relative importance or a specific number of the technical features indicated. Therefore, the features defined as "first", "second", etc. can explicitly or implicitly include at least one of the features. In the description of the present application, the meaning of "a plurality of" is at least two, such as two, three, etc., unless otherwise explicitly specified and limited.
[0041] In the present application, unless otherwise explicitly specified and limited, the terms "mounting", "connecting", "connecting", "fixing" and the like should be understood broadly, for example, it can be fixedly connected, or it can be detachably connected, or it can be integrated; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium; it can be the internal communication of two elements or the interaction relationship between two elements, unless otherwise explicitly limited. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.
[0042] In the present application, unless otherwise explicitly specified and limited, the first feature is "on" or "under" the second feature, which can be direct contact between the first and second features, or indirect contact between the first and second features through an intermediate medium. Moreover, the first feature "above", "over" and "on" the second feature can be directly above or obliquely above the first feature, or only indicate that the horizontal height of the first feature is higher than that of the second feature. The first feature "below", "above" and "under" the second feature can be directly above or obliquely above the first feature, or only indicate that the horizontal height of the first feature is less than that of the second feature.
[0043] In the description of the present application, the description of the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In the present application, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in the present application and the features of different embodiments or examples without contradiction.
[0044] As Figures 1-4 shown, the present application provides a technical solution: a fire prevention early warning method for a mountain photovoltaic power station, based on a multi-type sensor network deployed in the mountain photovoltaic power station, real-time acquisition of multi-dimensional monitoring data, the multi-dimensional monitoring data including temperature data, smoke concentration data, image data and environmental wind speed data;
[0045] In the mountain photovoltaic power station, a multi-type sensor network is constructed to obtain multi-dimensional monitoring data in real time. Distributed fiber temperature sensors are laid along the cables of the photovoltaic array to serve as monitoring nodes at every 10-meter interval, and real-time cable temperature data is collected at a sampling frequency of 1 time per second. Temperature abnormal changes caused by overload and poor contact of the cables are monitored.
[0046] Infrared smoke sensors are distributed in each area of the power station at a density of one per 50 square meters, and real-time monitoring of the smoke concentration in the environment is performed at a sampling frequency of 0.5 times per second, so that the smoke concentration can be sensed in a timely manner when it is relatively low.
[0047] High-definition monitoring cameras are installed at key positions and heights in the power station to ensure that the entire power station area can be covered. Image data is collected at a frame rate of 15 frames per second to identify fire-related features such as flames and smoke.
[0048] Wind speed and direction sensors are installed at an open and unobstructed area of the power station at a height of 5 meters to monitor the environmental wind speed and direction in real time at a sampling frequency of 2 times per second. This provides data support for evaluating the fire spread trend. Through the above-mentioned sensor network, multi-dimensional monitoring data including temperature data, smoke concentration data, image data, and environmental wind speed data are obtained.
[0049] The obtained multi-dimensional monitoring data is preprocessed to remove noise data and interpolate missing data.
[0050] For the obtained multi-dimensional monitoring data, wavelet transform denoising algorithm is used to process the temperature data and smoke concentration data. The specific wavelet transform formula is: wherein, is the original signal, is the wavelet coefficient, is the wavelet basis function. In practical applications, the Daubechies wavelet basis function (such as db4) is selected, the decomposition layer is set to 3 layers, the wavelet coefficients of different sizes and positions are calculated, then the soft threshold method is used to process the wavelet coefficients to remove noise components, and the signal is reconstructed by inverse wavelet transform to realize denoising of the temperature data and smoke concentration data. For missing data, cubic spline interpolation is used to complete the data. According to the known data points before and after the missing data, a cubic spline function is constructed to calculate the data value at the missing position, ensuring the integrity of the data. In addition, all data is normalized to convert different types and dimensions of data into a unified numerical range (such as the interval [0, 1]) for subsequent data fusion and analysis.
[0051] The preprocessed multi-dimensional monitoring data is input into a preset fusion analysis model, and the fusion analysis model is constructed based on a deep learning algorithm, specifically a convolutional neural network. First, the historical monitoring data of the power station in the past two years and the corresponding fire occurrence event data are collected, the historical monitoring data is feature extracted, including the temperature change trend, the mutation characteristics of the smoke concentration, the flame color and shape characteristics in the image, the wind speed size and direction change, etc., then the extracted feature data is divided into a training set, a validation set and a test set, the training set data is input into the CNN model for training, the weight coefficients of the model are continuously adjusted through the back propagation algorithm and the bias term b, so that the error between the prediction result of the model and the actual fire occurrence condition is minimized
[0052] , wherein F is the fused data value, D is the standard value of the i-th type of monitoring data, is the weight coefficient of the i-th type of monitoring data, b is the bias term, and n is the type of monitoring data, in this application n = 4, corresponding to temperature data, smoke concentration data, image data and environmental wind speed data. Through the formula, different types of monitoring data are fused according to their respective weights to obtain the fusion data of the fire risk condition.
[0053] Based on the fused data, combined with the topographic and geomorphic data and equipment operating state data of the mountain photovoltaic power station, the fire risk assessment algorithm is used for judgment. The topographic and geomorphic data includes altitude, slope, slope direction and vegetation coverage, which is obtained through geographic information system. The calculation formula of fire spread coefficient G is: , wherein k is an empirical coefficient (determined according to local historical fire data analysis, the value range is 0.5-1.5, and k = 1 in this embodiment), slope is the slope value, is the vegetation coverage (the value range is 0-1). The equipment operating state data includes the working voltage and working current of the photovoltaic module and the operating parameters of the inverter, which are collected in real time through the SCADA system. The calculation method of abnormal index E is: , wherein is the j-th parameter real-time monitoring value, is the historical normal operation mean value, is the historical standard deviation, is the parameter weight coefficient (determined by chromatography analysis method, the value range is 0.1-0.3 in this embodiment), and m is the number of equipment operating parameters (m = 6 in this embodiment), including the working voltage and working current of the photovoltaic module, the input voltage, input current, output voltage and output current of the inverter. The calculation formula of the fire risk assessment algorithm is:
[0054] Wherein, R is the fire risk value, T is the normalized value of temperature data, S is the normalized value of smoke concentration data, I is the fire feature value of image data, V is the influence factor of environmental wind speed data, G is the fire spread coefficient of topographic data, E is the abnormal index of equipment operation state data, 、 、 、 、 、 The weight coefficients of the corresponding data are trained and optimized by historical data (in this embodiment, = 0.3, = 0.2, = 0.2, = 0.1, = 0.1, = 0.1). Through the formula, the fire risk value is calculated by comprehensively considering various factors, so as to accurately evaluate the current fire risk level of the power station.
[0055] When it is judged that the fire risk value R exceeds the preset risk threshold (the preset risk threshold is 0.6 in this embodiment), a fire prevention warning signal is generated, and the specific position of the warning area is determined according to the area of the mountain photovoltaic power station (the power station is divided into 100 grid areas of 100m x 100m),
[0056] When it is judged that the fire risk value exceeds the preset risk threshold, a fire prevention warning signal is generated, and the specific position of the warning area is determined according to the area division of the mountain photovoltaic power station. After generating the fire prevention warning, according to the specific position of the warning area and the fire risk value, a preset emergency treatment scheme is called, if the fire risk value is between 0.6-0.7, the regional inspection is started, and the staff is arranged to check the warning area in detail, if the fire risk value is between 0.7-0.8, the sprinkler fire extinguishing system is started, and the non-key equipment power supply of the warning area is cut off, if the fire risk value is greater than 0.8, all power supplies of the warning area are immediately cut off, the whole station sprinkler fire extinguishing system is started, and the alarm information containing the position of the warning area, the fire risk value and other detailed information is sent to the monitoring center.
[0057] A fire prevention warning system for a mountain photovoltaic power station, comprising:
[0058] A data acquisition module for acquiring multi-dimensional monitoring data in real time through a multi-class sensor network;
[0059] The module obtains multi-dimensional monitoring data in real time through a multi-type sensor network. The module includes a sensor interface unit for communicating with distributed optical fiber sensors, infrared smoke sensors, high-definition monitoring cameras, and wind speed and direction sensors, supports multiple communication protocols such as Modbus, TCP / IP, and Zigbee, and can select an appropriate protocol for data transmission according to the characteristics of different sensors. The data acquisition unit can accurately control the data acquisition of each sensor according to the preset sampling frequency and acquisition strategy, ensuring the real-time and accuracy of the data. The data caching unit temporarily stores the acquired data to prevent data loss during transmission and serves as a data buffer to ensure the stability of data transmission. The data transmission unit uses a combination of wired and wireless transmission methods to quickly and stably transmit the acquired data to the data preprocessing module. Encryption technology is used during transmission to ensure data security.
[0060] The data preprocessing module is used to preprocess the acquired multi-dimensional monitoring data. The module receives the multi-dimensional monitoring data transmitted by the data acquisition module and preprocesses it. The preprocessing includes a noise removal unit that uses wavelet transform denoising algorithm to denoise temperature data and smoke concentration data, and uses image filtering algorithms such as median filtering and Gaussian filtering to denoise image data, to improve data quality. The missing value processing unit uses cubic spline interpolation, linear interpolation, and other methods to complete missing data, ensuring data integrity. The data normalization unit normalizes temperature data, smoke concentration data, and equipment operating parameters using the min-max normalization method, mapping data to the [0, 1] interval, so that different types of data have the same dimension and value range, facilitating subsequent data fusion and analysis.
[0061] The data fusion analysis module is used to input the preprocessed multi-dimensional monitoring data into the fusion analysis model for fusion processing. The module inputs the multi-dimensional monitoring data processed by the data preprocessing module into the preset fusion analysis model for fusion processing. The module includes a model storage unit for storing the trained CNN-based fusion analysis model, which can be updated and optimized according to actual needs. The feature extraction unit uses deep learning algorithms to extract features from the input multi-dimensional data, extracting key features that reflect fire risk. The data fusion calculation unit performs fusion calculation on the extracted feature data to obtain the fused data value. The model update unit can retrain and update the fusion analysis model based on new historical monitoring data and fire occurrence conditions, continuously improving the accuracy and adaptability of the model to better respond to different fire risk scenarios.
[0062] The model update unit can retrain and update the fusion analysis model based on new historical monitoring data and fire occurrence conditions, continuously improving the accuracy and adaptability of the model to better respond to different fire risk scenarios.
[0063] The risk assessment judgment module is used for fire risk assessment judgment based on the fused data combined with the topographic data and the equipment operation state data. Based on the fused data output by the data fusion analysis module, combined with the topographic data obtained from the GIS system and the equipment operation state data collected by the SCADA system, the fire risk assessment judgment is performed through a preset fire risk assessment algorithm. The module includes a topographic data processing unit, which analyzes and calculates the obtained data such as altitude, slope, slope direction and vegetation coverage, and calculates the fire risk value R according to the formula The fire risk value R is calculated by the risk calculation unit according to the fire risk assessment algorithm formula The abnormal index E is calculated by the equipment operation state data analysis unit, which processes the operation parameters of the photovoltaic components and the inverters according to the formula The fire risk value R is calculated by the risk calculation unit according to the fire risk assessment algorithm formula
[0064] The warning generation module is used for generating a fire prevention warning signal and determining the position of the warning area when the fire risk value exceeds the preset threshold. When the risk assessment judgment module judges that the fire risk value exceeds the preset threshold, the module generates a fire prevention warning signal, which includes a warning signal generation unit, generates different types of warning signals such as sound and light alarm signals, SMS alarm signals and email alarm signals according to the risk level, to meet the warning needs in different scenarios, a warning area positioning unit, which quickly and accurately determines the specific position of the warning area according to the regional division information of the power station and the position information of the sensor, and a warning information publishing unit, which sends the generated warning signal and the warning area position information to relevant personnel and departments through various ways such as display screen, SMS platform and email system, to ensure that the warning information can be timely conveyed.
[0065] The emergency processing module is further included for calling preset emergency disposal scheme according to the early warning area position and the fire risk value, and the module includes an emergency scheme storage unit for storing emergency processing schemes under different risk levels, such as starting a sprinkler fire extinguishing system, cutting off power, evacuating personnel, etc., a scheme calling and executing unit for automatically calling corresponding emergency processing scheme according to actual early warning situation and sending control instructions to related devices (such as a sprinkler system controller, a power switch controller, etc.) to execute emergency operation, and an emergency processing feedback unit for monitoring execution of the emergency processing measures in real time and feeding back execution results to the monitoring center and related personnel to timely adjust the emergency processing strategy and ensure effectiveness of the emergency processing. In addition, the data acquisition module, the data preprocessing module, the data fusion analysis module, the risk assessment and judgment module, and the early warning generation module are integrated in the central processing unit, the central processing unit adopts a high-performance embedded processor (such as an ARM Cortex-A72 architecture processor) and has strong data processing and operation capacity to ensure real-time performance and stability of the system, and has the advantages of low power consumption, small size, etc., and is suitable for long-term operation in the complex environment of the mountain photovoltaic power station.
[0066] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, replacements and changes can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended embodiments and their equivalents.
Claims
1. A fire prevention early warning method for a mountain photovoltaic power station, characterized in that, The method comprises the following steps: Real-time acquisition of multi-dimensional monitoring data including temperature data, smoke concentration data, image data and environmental wind speed data based on a multi-type sensor network deployed in a mountain photovoltaic power station; Preprocessing of the acquired multi-dimensional monitoring data to remove noise data and interpolate missing data; Inputting the preprocessed multi-dimensional monitoring data into a preset fusion analysis model constructed based on a deep learning algorithm for fusion processing of the data according to the following formula: wherein F is the fused data value, D is the standard value of the i-th type of monitoring data, is the weight coefficient of the i-th type of monitoring data, b is the bias term, and n is the type of monitoring data. Based on the fused data, combined with the topographic and geomorphic data and equipment operation state data of the mountain photovoltaic power station, a preset fire risk assessment algorithm is used for judgment, and the calculation formula of the fire risk assessment algorithm is: ; Wherein, R is the fire risk value, T is the normalized value of temperature data, S is the normalized value of smoke concentration data, I is the fire feature value of image data, V is the influence factor of environmental wind speed data, G is the fire spread coefficient of topographic data, E is the abnormal index of equipment operation state data, is the weight coefficient corresponding to the data; When it is judged that the fire risk value exceeds a preset risk threshold, a fire prevention warning signal is generated, and the specific location of the warning area is determined according to the regional division of the mountain photovoltaic power station.
2. The fire-prevention early warning method for a mountain photovoltaic power station according to claim 1, characterized in that: The multi-type sensor network comprises a distributed optical fiber temperature sensor, an infrared smoke sensor, a high-definition monitoring camera and a wind speed and direction sensor, wherein the distributed optical fiber temperature sensor is arranged along the cable of the photovoltaic array for monitoring the cable temperature.
3. The fire-prevention early warning method for a mountain photovoltaic power station according to claim 1, characterized in that: In the preprocessing step of the multi-dimensional monitoring data, a wavelet transform denoising algorithm is used for denoising processing of the temperature data and the smoke concentration data, and the wavelet transform formula is: wherein, is the original signal, is the wavelet coefficient, is the wavelet basis function.
4. The fire-prevention early warning method for a mountain photovoltaic power station according to claim 1, characterized in that: The construction process of the fusion analysis model comprises: collecting historical monitoring data and corresponding fire time data, performing feature extraction on the historical monitoring data, learning the features by using a convolutional neural network, and optimizing weight coefficients by using a back propagation algorithm and a bias term b.
5. The method according to claim 1, characterized in that: The topographic and geomorphic data includes elevation, slope, aspect and vegetation coverage, which are obtained by a geographic information system and used to calculate the fire spread coefficient G, and the calculation formula of the fire spread coefficient is: wherein K is an empirical coefficient, slope is the slope value, is the vegetation coverage.
6. The fire-prevention early warning method for a mountain photovoltaic power station according to claim 1, characterized in that: The equipment operating state data includes the working voltage and working current of the photovoltaic module and the operating parameters of the inverter, which are collected in real time by a SCADA system, and the calculation method of the anomaly index E is: wherein, is a real-time value of the jth equipment operating parameter, is a historical average value of the jth equipment operating parameter, is a historical standard deviation of the jth equipment operating parameter, is an importance coefficient of the jth equipment operating parameter, and m is the number of equipment operating parameters.
7. A fire prevention and early warning method for a mountain photovoltaic power station according to claim 1, characterized in that: After the fire prevention warning is generated, a preset emergency handling scheme is called according to the specific location of the warning area and the fire risk value, and the emergency handling scheme includes starting the sprinkler fire extinguishing system, cutting off the power supply of the corresponding area and sending alarm information to the monitoring center.
8. A fire prevention early warning system for a mountain photovoltaic power station, characterized in that, It comprises: A data acquisition module for acquiring multi-dimensional monitoring data in real time through a multi-type sensor network; A data preprocessing module for preprocessing the acquired multi-dimensional monitoring data; A data fusion analysis module for inputting the preprocessed multi-dimensional monitoring data into a fusion analysis model for fusion processing; A risk assessment and judgment module for fire risk assessment and judgment according to the fused data combined with topographic and geomorphic data and equipment operating state data; A warning generation module for generating a fire prevention warning signal when the fire risk value exceeds a preset threshold and determining the location of the warning area.
9. The fire-prevention early-warning system of a mountain photovoltaic power station according to claim 8, characterized in that, The emergency handling module is also included for calling a preset emergency handling scheme according to the location of the warning area and the fire risk value.
10. The fire-prevention early warning system of a mountain photovoltaic power station according to claim 8, characterized in that, The data acquisition module, the data preprocessing module, the data fusion analysis module, the risk assessment and judgment module and the warning generation module are integrated in a central processing unit, and the central processing unit adopts an embedded processor.