Photovoltaic integrated fire linkage early warning method and system for building

Abstract

The invention provides an integrated fire linkage early warning method and system for building photovoltaics, and relates to the technical field of fire early warning, and the method comprises the steps: obtaining and setting plane distribution information; dividing a plurality of photovoltaic string linkage areas; setting a photovoltaic module monitoring point, and setting a single-point monitoring sensor; component abnormal information is determined; setting a plurality of transmission equipment linkage nodes; performing anomaly analysis according to the monitoring data of the plurality of transmission linkage nodes, and determining transmission equipment anomaly information; and performing fire probability prediction according to the assembly abnormal information and the transmission equipment abnormal information, and performing early warning based on a fire probability prediction result, thereby solving the problems that in the prior art, since unified sensors are mostly arranged at different positions of a building photovoltaic system to perform fire monitoring and early warning, the arrangement positions of the sensors are inaccurate, and the cost is low. The technical problems are solved, and the technical effects of improving the abnormity monitoring precision and efficiency and further improving the fire hazard early warning accuracy are achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of fire warning, in particular to an integrated fire linkage warning method and system for building photovoltaics. BACKGROUND

[0002] In recent years, with the decline of system cost, the degree of electricity cost of solar photovoltaic power generation technology has been far lower than that of conventional fossil fuel power generation cost, and therefore it has been widely used. However, with the wide application of building photovoltaics, photovoltaic power station fire incidents have become an important factor threatening people and building safety. In the prior art, a unified sensor is mostly set at different positions of the photovoltaic system to monitor and warn fire, but the setting position of the sensor is not accurate, which leads to inaccurate monitoring results.

[0003] At present, in the prior art, there is a technical problem that the fire monitoring and warning results are inaccurate because a unified sensor is mostly set at different positions of the photovoltaic system to monitor and warn fire, and the setting position of the sensor is not accurate. SUMMARY

[0004] The present application provides an integrated fire linkage warning method and system for building photovoltaics to solve the technical problem that the fire monitoring and warning results are inaccurate because a unified sensor is mostly set at different positions of the photovoltaic system to monitor and warn fire, and the setting position of the sensor is not accurate.

[0005] According to a first aspect of the present application, an integrated fire linkage warning method for building photovoltaics is provided, comprising: collecting setting information of N photovoltaic strings of building photovoltaics, and obtaining setting plane distribution information; dividing a plurality of photovoltaic string linkage areas according to the setting plane distribution information, and setting a regional monitoring sensor for each photovoltaic string linkage area; determining an abnormal probability component based on historical record data of photovoltaic components of the N photovoltaic strings, setting a photovoltaic component monitoring point, and setting a single-point monitoring sensor for the photovoltaic component monitoring point; performing abnormal analysis according to monitoring data of the single-point monitoring sensor and the regional monitoring sensor, and determining component abnormal information; obtaining a transmission path node device of building photovoltaics, and setting a plurality of transmission device linkage nodes; performing abnormal analysis according to monitoring data of the plurality of transmission linkage nodes, and determining transmission device abnormal information; performing fire probability prediction according to the component abnormal information and the transmission device abnormal information, and performing warning based on the fire probability prediction result.

[0006] According to a second aspect of this application, an integrated fire linkage early warning system for building-integrated photovoltaics (BIPV) is provided, comprising: a photovoltaic string distribution information acquisition module, which is used to collect setting information of N photovoltaic strings of the building-integrated photovoltaic system and obtain setting planar distribution information; a linkage area division module, which is used to divide multiple photovoltaic string linkage areas according to the setting planar distribution information, and each photovoltaic string linkage area is equipped with an area monitoring sensor; and a monitoring point setting module, which is used to determine abnormal probability components based on historical data of photovoltaic modules of the N photovoltaic strings, set photovoltaic module monitoring points, and set single-point monitoring for the photovoltaic module monitoring points. The system includes: a monitoring sensor; a first anomaly analysis module, which performs anomaly analysis based on monitoring data from the single-point monitoring sensor and the area monitoring sensor to determine component anomaly information; a linkage node setting module, which acquires the transmission path node devices of the building photovoltaic system and sets multiple transmission device linkage nodes; a second anomaly analysis module, which performs anomaly analysis based on monitoring data from multiple transmission linkage nodes to determine transmission device anomaly information; and a fire probability prediction module, which predicts the fire probability based on the component anomaly information and the transmission device anomaly information, and issues an early warning based on the fire probability prediction results.

[0007] The beneficial effects that can be achieved by adopting one or more technical solutions in this application are as follows:

[0008] This system collects the setup information of N photovoltaic (PV) strings in a building-integrated photovoltaic (BIPV) system, obtaining their planar distribution information. Based on this distribution, multiple PV string linkage zones are defined, and a zone monitoring sensor is installed in each zone. Based on historical data of the PV modules in the N strings, modules with abnormal probability are identified, and monitoring points are set for each module. Single-point monitoring sensors are then installed at these monitoring points. Anomaly analysis is performed based on the monitoring data from both single-point and zone sensors to determine module anomalies. The system also acquires information on the transmission path nodes of the BIPV system, setting up multiple linkage nodes. Anomaly analysis is performed on the monitoring data from these linkage nodes to determine transmission equipment anomalies. Finally, fire probability prediction is performed based on the module and transmission equipment anomalies, and an early warning is issued based on the prediction results. This combined monitoring by single-point and zone sensors improves the accuracy and efficiency of anomaly detection, thereby enhancing the accuracy of fire early warning systems. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in this application or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. The accompanying drawings, which constitute a part of this application, are used to provide a further understanding of this application. The illustrative embodiments of this application and their descriptions are used to explain this application and do not constitute an improper limitation of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0010] Figure 1 A schematic flowchart illustrating an integrated fire-linked early warning method for building photovoltaic systems provided in this application embodiment;

[0011] Figure 2 This is a schematic diagram of an integrated fire alarm system for building photovoltaics provided in an embodiment of this application.

[0012] Explanation of reference numerals in the attached diagram: 11 Photovoltaic string distribution information acquisition module, 12 linkage area division module, 13 monitoring point setting module, 14 first anomaly analysis module, 15 linkage node setting module, 16 second anomaly analysis module, and 17 fire probability prediction module. Detailed Implementation

[0013] To make the objectives, technical solutions, and advantages of this application more apparent, exemplary embodiments of this application will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments of this application. It should be understood that this application is not limited to the exemplary embodiments described herein.

[0014] The terminology used in this specification is for describing embodiments and not for limiting the application. As used in the specification, the singular terms “a,” “an,” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. When used in the specification, the terms “comprising” and / or “including” specify the presence of a step, operation, element, and / or component, but do not preclude the presence or addition of one or more other steps, operations, elements, components, and / or groups thereof.

[0015] Unless otherwise defined, all terms used in this specification (including technical and scientific terms) shall have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. Terms, as defined in common dictionaries, shall not be interpreted in an idealized or overly formal sense unless expressly defined herein. Throughout this specification, the same reference numerals denote the same elements.

[0016] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for display, data used for analysis, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0017] Example 1

[0018] Figure 1 A diagram illustrating an integrated fire alarm and early warning method for building photovoltaic systems provided in this application embodiment, the method comprising:

[0019] Collect the configuration information of N photovoltaic strings in building-integrated photovoltaics (BIPV) systems and obtain their planar distribution information.

[0020] A photovoltaic string refers to a circuit unit with a certain DC power output formed by connecting several photovoltaic modules in series in a building-integrated photovoltaic (BIPV) power generation system. This involves collecting the configuration information of N photovoltaic strings based on actual conditions, where N is an integer greater than 0. The planar distribution information refers to the distribution location of the N photovoltaic strings. During the initial design of a BIPV power generation system, professionals in the field determine the installation locations of the photovoltaic strings based on the building's structure before installation. The method provided in this application embodiment is used for fire-linked early warning of BIPV systems already in use; therefore, the planar distribution information of the photovoltaic strings has already been configured and can be collected and uploaded by the user.

[0021] Based on the aforementioned planar distribution information, multiple photovoltaic array series-connected operation areas are divided, and a regional monitoring sensor is installed in each photovoltaic array series-connected operation area.

[0022] In a preferred embodiment, it further includes:

[0023] The sensing parameters of the regional monitoring sensors are acquired, including sensing range and sensing constraint space parameters; based on the set planar distribution information, the distribution area and distribution building structure are determined; according to the sensing range and sensing constraint space parameters, interval division constraints are set, and the set planar distribution information is divided into regions to obtain multiple photovoltaic array series linkage regions.

[0024] Based on the aforementioned planar distribution information, multiple photovoltaic string connection zones are divided. Each photovoltaic string connection zone is equipped with a zone monitoring sensor. The zone monitoring sensor refers to a sensor used for monitoring photovoltaic string anomalies, such as a temperature sensor, a light sensor, or a gas sensor. The sensor can be selected according to the actual situation, and there are no restrictions on it.

[0025] The specific process of dividing multiple photovoltaic (PV) string interconnection areas based on the aforementioned planar distribution information is as follows: First, obtain the sensing parameters of the area monitoring sensors. These parameters include the sensing range and sensing constraint space parameters. The sensing range refers to the range of indicators that the area monitoring sensors can monitor, such as a temperature sensor that can monitor a temperature range of -20℃ to 30℃. The sensing constraint space parameters refer to the spatial range that the area monitoring sensors can monitor, such as monitoring PV strings within 1 square meter. Both the sensing range and the sensing constraint space parameters can be directly read from the user manual of the area monitoring sensors. Further, based on the aforementioned planar distribution information, determine the distribution area and the distributed building structure. The distribution area is the total distribution area of ​​the N PV strings, and the distributed building structure refers to the distribution location of the N PV strings on a building, such as the roof, building side, etc., which needs to be determined according to the actual situation. Then, based on the sensing range and sensing constraint space parameters, interval division constraints are set. Simply put, the size of a monitoring area is determined by the sensing range and sensing constraint space parameters, ensuring that the monitoring area meets these requirements. This monitoring area serves as the interval division constraint condition, and the planar distribution information is divided into multiple regions of the same size as the monitoring area. These regions are then used as multiple photovoltaic module series-connected operation regions. Each photovoltaic module series-connected operation region has a region monitoring sensor for anomaly monitoring. Thus, the division of photovoltaic module series-connected operation regions based on the sensing range and sensing constraint space parameters of the region monitoring sensors provides a foundation for subsequent component anomaly analysis.

[0026] Based on the historical data of photovoltaic modules in N photovoltaic strings, identify modules with abnormal probability, set monitoring points for photovoltaic modules, and set single-point monitoring sensors for the monitoring points of photovoltaic modules.

[0027] In a preferred embodiment, it further includes:

[0028] The system acquires the usage time, deployment environment, and anomaly records of photovoltaic modules; performs anomaly evaluation based on the usage time, deployment environment, and anomaly records respectively, and determines the anomaly probability; based on the anomaly probability, it determines the anomaly probability component, wherein the anomaly probability component is a photovoltaic module whose anomaly probability reaches a preset probability threshold.

[0029] Based on historical data of photovoltaic modules in N photovoltaic strings, which refers to the usage records of the photovoltaic modules in the N photovoltaic strings over a past period, including usage duration, deployment environment, and anomaly records, this data can be retrieved by those skilled in the art. Further, components with a high probability of failure are identified. Monitoring points are set for these components, and single-point monitoring sensors are installed at each monitoring point. The monitoring accuracy of single-point sensors is greater than that of area monitoring sensors.

[0030] The process of determining the probability of anomalies based on historical data of photovoltaic modules in N photovoltaic strings is as follows: The historical data is extracted to include the usage duration (the duration of the photovoltaic module from installation to the present, or the duration since replacement if it has been replaced), the installation environment (humidity parameters of the photovoltaic module's location), and anomaly records (historical anomalies, anomaly times, anomaly frequencies, etc.). Anomalies are evaluated based on the usage duration, installation environment, and anomaly records to determine the anomaly probability. Generally, photovoltaic modules have a lifespan and need to be replaced after reaching that lifespan. The closer the usage duration is to the lifespan, the higher the anomaly probability. The more humid the installation environment, the higher the probability of anomalies. The higher the number of anomalies recorded, the higher the probability of subsequent anomalies. Specifically, those skilled in the art can retrieve usage duration samples, deployment environment samples, and anomaly record samples of similar photovoltaic modules, and obtain anomaly probability samples of similar photovoltaic modules based on existing technologies. Using the usage duration samples, deployment environment samples, and anomaly record samples as input parameters, and the anomaly probability samples as output parameters, an anomaly evaluation model can be trained based on existing machine learning models. Machine model training is a common technique used by those skilled in the art and will not be elaborated upon here. Then, the usage duration, deployment environment, and anomaly records of the photovoltaic modules are input into the anomaly evaluation model to output the anomaly probability.

[0031] Based on the aforementioned anomaly probability, an anomaly probability component is determined. This component refers to photovoltaic modules whose anomaly probability reaches a preset probability threshold. The preset probability threshold is set by those skilled in the art based on actual conditions. In other words, photovoltaic modules with an anomaly probability reaching the preset threshold have a higher probability of malfunctioning, and are therefore designated as monitoring points for these modules. A single-point monitoring sensor is installed at each of these locations. The indicators monitored by the single-point monitoring sensor are the same as those monitored by the area monitoring sensor, but the single-point monitoring sensor has higher monitoring accuracy. It is used to monitor the anomaly probability component. By setting up a single-point monitoring sensor, it is easier to conduct joint monitoring with the area monitoring sensor subsequently, improving the accuracy of anomaly analysis.

[0032] Based on the monitoring data from the single-point monitoring sensor and the area monitoring sensor, anomaly analysis is performed to determine component anomaly information;

[0033] In a preferred embodiment, it further includes:

[0034] The system activates the regional monitoring sensors to monitor the photovoltaic module series linkage area and acquires monitoring data for each linkage area. Based on the positional relationship of each linkage area and the monitoring data of each linkage area, the system calculates the difference value and uses the difference value to locate anomalies and determine the abnormal linkage area. Based on the abnormal linkage area, the system locates the single-point monitoring sensor of the linkage area to monitor the photovoltaic module monitoring point and acquires the monitoring data of the photovoltaic module monitoring point. The system determines whether there is an anomaly in the monitoring data of the photovoltaic module monitoring point. If there is, the system determines the module anomaly information based on the monitoring data of the photovoltaic module monitoring point. If there is no anomaly, the system determines the module anomaly information based on the abnormal linkage area.

[0035] In a preferred embodiment, it further includes:

[0036] The image acquisition device is activated to acquire images based on the abnormal linkage area and obtain regional image information; the regional image information is segmented according to a preset segmentation region; the segmented image is traversed using a preset abnormal feature convolution kernel to locate the abnormal photovoltaic module and determine the abnormal information of the module.

[0037] Based on the monitoring data from the single-point monitoring sensor and the area monitoring sensor, anomaly analysis is performed to determine the abnormal information of the components. In simple terms, the abnormal linkage area is first located by using the monitoring data from the area monitoring sensor, and then the photovoltaic components in the abnormal linkage area are monitored at a single point by the single-point monitoring sensor to determine the abnormal information of the components. The specific process is detailed below.

[0038] Specifically, the regional monitoring sensors are activated to monitor the interconnected areas of the photovoltaic modules, acquiring monitoring data for each interconnected area. This data is the sensor data output by the regional monitoring sensors, such as the temperature of each interconnected area output by a temperature sensor. Based on the positional relationship of each interconnected area and the monitoring data of each interconnected area, a difference value calculation is performed. This involves calculating the difference between the monitoring data of any two adjacent interconnected areas to obtain the difference value between them. Then, anomaly localization is performed based on this difference value. Specifically, a preset difference value threshold can be determined by those skilled in the art. This preset difference value threshold represents the allowable range of differences when the photovoltaic modules are not abnormal. It can be set by those skilled in the art based on practical experience. Interconnected areas with difference values ​​greater than the preset difference value threshold are identified as abnormal interconnected areas. It should be noted that when monitoring is carried out through regional monitoring sensors, continuous monitoring data analysis can be performed. That is to say, the monitoring data of a certain area may not be accurate at a certain moment, for example, due to external environmental factors causing data anomalies. Therefore, the monitoring data can be compared and analyzed over a continuous period of time. If the monitoring data is abnormal over a continuous period of time, the abnormal linkage area can be determined based on the monitoring data; conversely, if the monitoring data returns to normal over a continuous period of time, it indicates that the area is a normal area.

[0039] Based on the abnormal linkage area, a single-point monitoring sensor in that linkage area is located to monitor the photovoltaic module monitoring point, and the sensor data output by the single-point monitoring sensor is used as the monitoring data of the photovoltaic module monitoring point. Further, it is determined whether the monitoring data of the photovoltaic module monitoring point is abnormal. Specifically, based on historical experience, normal monitoring data corresponding to the photovoltaic module monitoring point can be obtained, i.e., monitoring data when no abnormalities occur. The monitoring data of the photovoltaic module monitoring point is then compared with the normal monitoring data to determine whether the monitoring data of the photovoltaic module monitoring point is within the normal monitoring data range. If not, it indicates that the monitoring data of the photovoltaic module monitoring point is abnormal, and the abnormal information of the module is determined based on the monitoring data. Similarly, when monitoring through a single-point monitoring sensor, continuous monitoring data analysis can also be performed. If the monitoring data is abnormal for a continuous period of time, the abnormal information of the module is determined based on the monitoring data; conversely, if the monitoring data returns to normal for a continuous period of time, it indicates that the corresponding photovoltaic module monitoring point is operating normally.

[0040] The abnormal information of the components includes the monitoring points of the photovoltaic modules and the corresponding monitoring data, which facilitates subsequent fire probability analysis.

[0041] When no anomaly is found, the anomaly information of the component is determined based on the anomaly linkage area. The specific process is as follows:

[0042] The image acquisition device is activated. This device refers to an existing visual sensor used for image acquisition, such as a smart camera, infrared camera, etc., depending on the specific circumstances. The image acquisition device is used to acquire images of the abnormal linkage area, obtaining regional image information. The regional image information is then segmented according to a preset segmentation region. This preset segmentation region is set by those skilled in the art and corresponds to a preset abnormal feature convolution kernel. The preset abnormal feature convolution kernel refers to the convolution kernel of the preset segmentation region under normal conditions. Specifically, it can be determined by those skilled in the art based on historical experience, i.e., by collecting images of the preset segmentation region under various abnormal conditions over a historical period and calculating the convolution kernel to obtain the preset abnormal feature convolution kernel. Then, image segmentation is performed on the regional image information based on the preset segmentation region to obtain segmented images. The convolution kernel corresponding to the segmented image is calculated and compared with the preset abnormal feature convolution kernel. If the convolution kernel corresponding to the segmented image is the same as the preset abnormal feature convolution kernel, the region where the segmented image is located is considered an abnormal photovoltaic module, and the corresponding abnormal situation is matched according to the corresponding preset abnormal feature convolution kernel as the module's abnormal information.

[0043] Therefore, by using regional monitoring sensors and single-point monitoring sensors to determine component anomaly information, the accuracy of anomaly analysis is improved, thereby enhancing the accuracy of fire linkage early warning.

[0044] Obtain the transmission path node devices of building-integrated photovoltaics (BIPV) and set up multiple transmission device linkage nodes;

[0045] Current photovoltaic string technology typically uses around 20 modules arranged in a string. A 1000V or 1500V DC high-voltage line enters a DC combiner box, then a centralized inverter or string inverter for DC-to-AC conversion. The AC power is then fed into the power grid. The transmission path nodes, such as the DC combiner box, centralized inverter, and string inverter, are the equipment responsible for transmitting and converting electrical energy. The specific details need to be determined based on the actual conditions of building-integrated photovoltaics (BIPV). Alternatively, multiple interconnected transmission path nodes can be set up, where each device in the transmission path node acts as an interconnected node for multiple transmission devices according to the sequence of power transmission.

[0046] Anomaly analysis is performed based on monitoring data from multiple transmission linkage nodes to determine abnormal information of transmission equipment;

[0047] Anomaly analysis is performed based on monitoring data from multiple transmission linkage nodes. This monitoring data refers to the monitoring information of the transmission path node devices corresponding to the multiple transmission linkage nodes. Specifically, it can be information such as voltage and current collected by electrical sensors, or temperature information collected by temperature sensors. Those skilled in the art can set up corresponding sensors at the multiple transmission linkage nodes to obtain the monitoring data. Then, based on existing technology, normal monitoring data from the multiple transmission linkage nodes is obtained. By comparing the monitoring data from the multiple transmission linkage nodes with the normal monitoring data, monitoring data that differs from the normal monitoring data is identified as anomaly information of the transmission equipment.

[0048] Fire probability is predicted based on the abnormal information of the components and transmission equipment, and an early warning is issued based on the fire probability prediction results.

[0049] In a preferred embodiment, it further includes:

[0050] Based on fire accident records, the logical influence coefficients of photovoltaic module monitoring data, transmission equipment monitoring data, and fire accidents are determined. Based on these logical influence coefficients, probability calculations are performed on the abnormal information of the modules and the abnormal information of the transmission equipment to determine the fire probability. When the fire probability of either the abnormal information of the modules or the abnormal information of the transmission equipment reaches a preset threshold, an automatic fire alarm system is connected, and a photovoltaic abnormal fire probability warning is sent. According to the transmission path of the building photovoltaic system, a transmission influence data chain is constructed, using the fire probability as the influence coefficient of the path nodes. Based on the fire accident records, the correlation influence of each transmission path node is obtained, and this correlation influence is added as an adjustment coefficient to the influence coefficient. Probability prediction is performed on each transmission path node, and when the preset threshold is reached, the photovoltaic abnormal fire probability warning is sent.

[0051] Fire probability is predicted based on the abnormal information of the components and transmission equipment, and an early warning is issued based on the fire probability prediction results. The specific process is as follows:

[0052] Based on fire accident records, the logical influence coefficients between photovoltaic module monitoring data, transmission equipment monitoring data, and fire accidents are determined. Fire accident records refer to historical fire accident records corresponding to building-integrated photovoltaic (BIPV) modules or photovoltaic modules of the same model, which can be accessed by those skilled in the art. Fire accident records include records of fire causes (e.g., photovoltaic module malfunction, transmission equipment malfunction). Based on this, existing correlation analysis methods, such as grey relational analysis (correlation analysis is a commonly used technique by those skilled in the art, and therefore will not be elaborated upon here), are used to analyze the fire accident records, obtaining the correlation coefficients between photovoltaic module malfunctions, transmission equipment malfunctions, and fire accidents, respectively, as the logical influence coefficients between photovoltaic module monitoring data, transmission equipment monitoring data, and fire accidents.

[0053] The probability of a fire is determined by calculating the probabilities of the abnormal information of the photovoltaic modules and the abnormal information of the transmission equipment based on the logical influence coefficients. Simply put, the degree of abnormality of the abnormal information of the modules and the transmission equipment is first calculated, i.e., the percentage difference between the abnormal information and the normal monitoring data. Then, the logical influence coefficients of the photovoltaic module monitoring data, the transmission equipment monitoring data, and the fire accident are multiplied by the percentage differences of the abnormal information of the modules and the transmission equipment, respectively, to obtain the fire probabilities corresponding to the abnormal information of the modules and the transmission equipment. When the fire probability of either the abnormal information of the modules or the abnormal information of the transmission equipment reaches a preset threshold, the automatic fire alarm system is connected, and a photovoltaic abnormal fire probability warning is sent. The preset threshold is set by professionals in the field. The automatic fire alarm system consists of a triggering device and a fire alarm device. When the fire probability of either the abnormal information of the modules or the abnormal information of the transmission equipment reaches the preset threshold, the triggering device triggers the fire alarm device, which alarms in the form of sound or light, reminding people in the building to take emergency evacuation, and simultaneously reminding relevant personnel to inspect the photovoltaic modules.

[0054] It should be noted that single-point monitoring sensors and area monitoring sensors can be sensing devices used to monitor different types of data, such as temperature, voltage, and current. Multiple sensing devices for monitoring different types of data can be set up at one location as single-point monitoring sensors and area monitoring sensors. Anomaly analysis is performed based on different types of data. Therefore, the component anomaly information and transmission device anomaly information obtained will also include at least one type of anomaly information, such as temperature anomaly, voltage anomaly, and current anomaly. Furthermore, the more types of anomalies are included in the final obtained component and transmission equipment anomaly information, the higher the anomaly level of the corresponding component and the higher the probability of a fire. Therefore, graded early warnings can be implemented based on the types of anomalies included in the component and transmission equipment anomaly information. For example, if there is only one type of data anomaly, a low-level warning is issued, without sound or light warnings, but simply by sending the component and transmission equipment anomaly information to the building photovoltaic maintenance management personnel. Subsequently, the maintenance management personnel can assign maintenance personnel to troubleshoot the fault, thereby reducing the waste of fire-fighting resources and improving the accuracy of the warning. Similarly, if there are multiple types of data anomalies, a high-level warning is activated, i.e., a warning is issued through sound or light. The specific warning level classification can be based on the actual type and number of sensors used, and there are no restrictions on this.

[0055] Meanwhile, buildings are generally equipped with existing fire early warning systems. These systems are used to issue fire alarms and control the evacuation of personnel after a fire occurs in the building itself. Therefore, when the triggering device activates the fire alarm device, it can connect to the existing fire early warning system and exchange component abnormality information and transmit equipment abnormality information with the existing fire early warning system. This facilitates personnel evacuation and fire alarm activation. At the same time, when the existing fire early warning system issues an early warning, it will also synchronously exchange the warning information with the integrated fire linkage early warning system for building photovoltaics provided in this application. This assists in the cut-off control of equipment in the building photovoltaic system, preventing the fire from linking the building itself with the building photovoltaic equipment and causing further damage.

[0056] Based on the transmission path of building-integrated photovoltaics (BIPV), a transmission impact data chain is constructed. In simpler terms, following the connection sequence of transmission path node devices, such as the path from the DC combiner box to the centralized inverter, each transmission path node device is considered a path node, and its corresponding fire probability is used as its impact coefficient. Further, based on the fire accident records, the correlation impact of each transmission path node is obtained. Specifically, the abnormal situations of each transmission path node in each fire record can be retrieved from the fire accident records. Then, based on existing correlation analysis methods (such as grey relational analysis), correlation analysis is performed on each transmission path node to obtain the corresponding correlation impact. This correlation impact is added as an adjustment coefficient to the corresponding logical impact coefficient, essentially superimposing the correlation impact and the original logical impact coefficient. This allows for probability prediction of each transmission path node. When the preset threshold is reached, a BIPV abnormal fire probability warning is sent, thereby achieving fire linkage early warning for building-integrated photovoltaics and improving the accuracy of the warning.

[0057] In a preferred embodiment, it further includes:

[0058] Connect to the automatic fire alarm system to obtain fire alarm information; calculate the building coverage rate based on the set planar distribution information, and divide the building into zones based on the building photovoltaic coverage rate; determine the fire range information based on the fire alarm information; determine the association probability based on the fire range information, building zones, and building photovoltaic coverage rate; when the association probability reaches a preset range, generate association cut-off control information.

[0059] The system connects to an automatic fire alarm system and receives fire alarm information, including the locations of photovoltaic (PV) modules where the fire probability reaches a preset threshold. Based on the set planar distribution information, the building coverage rate is calculated, and the building is divided into zones based on the PV coverage rate. The PV coverage rate refers to the degree of coverage of a building by a building photovoltaic (PV) system, reflecting the distribution of the PV system on the building, i.e., the proportion of the coverage area of ​​PV devices in different locations to each area. Zones are formed by grouping adjacent locations with similar or identical coverage areas. The fire range information is determined based on the fire alarm information; the fire range information is the area comprised of the locations of PV modules where the fire probability reaches the preset threshold. Based on the fire range information, building zones, and PV coverage rate, the correlation probability is determined. Simply put, a fire in one location may also cause a fire in a normally operating PV module in an adjacent location. Therefore, based on the fire range information, building zones, and PV coverage rate, the probability that the fire range information will also cause a fire in other zones within the building zone is determined. Specifically, different building zones have varying distances from fire-prone areas, resulting in different fire risks; the closer the distance, the higher the correlation probability. Building photovoltaic (PV) coverage also affects fire risk; higher coverage means more PV modules are exposed to fire, potentially increasing the correlation probability. Considering these factors, the correlation probability can be calculated using existing technology by acquiring historical data and training a machine learning model. When the correlation probability reaches a preset range (set by those skilled in the art based on practical experience), correlation cutoff control information is generated. This cuts off the circuitry of the PV modules within the building zone corresponding to the correlation probability, preventing fire migration.

[0060] In a preferred embodiment, it further includes:

[0061] The system acquires the periodic characteristics of building illumination; analyzes the abnormal impact of the periodic characteristics on the photovoltaic (PV) array series-moving area; sets a multi-level response hierarchy for area sensors and / or single-point monitoring sensors based on the abnormal impact relationship between the periodic characteristics and the PV array series-moving area; performs graded monitoring of the PV array series-moving area according to the multi-level response hierarchy; and replaces the multi-level response hierarchy of area sensors and / or single-point monitoring sensors when the periodic characteristics change, based on the abnormal impact relationship between the periodic characteristics and the PV array series-moving area.

[0062] The periodicity of building illumination is obtained, which refers to the variation in light received by building photovoltaic (PV) systems, such as the change curve of light intensity at different locations over a period of time. Based on this periodicity, the abnormal impact analysis of PV string fluctuation areas is performed. That is, the probability of abnormalities in PV strings varies with different illumination levels; for example, the probability of abnormalities is higher in areas with more drastic changes in light intensity or in more humid environments. Therefore, multiple PV string fluctuation areas can be sequentially arranged according to the degree of drastic change in light intensity or the order of humidity from high to low within the periodicity of building illumination. This yields the abnormal impact relationship between the periodicity and the PV string fluctuation areas. The earlier a PV string fluctuation area is ranked, the stronger the abnormal impact relationship between the periodicity and the PV string fluctuation area. Based on the anomalous impact relationship between the periodic characteristics and the photovoltaic (PV) array series-movement area, a multi-level response hierarchy of area sensors and / or single-point monitoring sensors is established. That is, the area with the strongest anomalous impact relationship is designated as the first level, allowing for the maximum number of area and single-point monitoring sensors. Similarly, for areas with weaker anomalous impact relationships, fewer area and single-point monitoring sensors are used, thus obtaining the multi-level response hierarchy for graded monitoring of the PV array series-movement area, using different numbers of sensors. When the periodic characteristics change, the multi-level response hierarchy of area sensors and / or single-point monitoring sensors is changed based on the anomalous impact relationship of the periodic characteristics and the PV array series-movement area. This involves updating the periodic characteristics, re-analyzing the anomalous impact of the PV array series-movement area, and setting the multi-level response hierarchy of area sensors and / or single-point monitoring sensors, thereby improving the accuracy of anomaly monitoring while reducing the number of monitoring sensors required.

[0063] Based on the above analysis, the beneficial effects that one or more technical solutions provided in this application can achieve are as follows:

[0064] This system collects the setup information of N photovoltaic (PV) strings in a building-integrated photovoltaic (BIPV) system, obtaining their planar distribution information. Based on this distribution, multiple PV string linkage zones are defined, and a zone monitoring sensor is installed in each zone. Based on historical data of the PV modules in the N strings, modules with abnormal probability are identified, and monitoring points are set for each module. Single-point monitoring sensors are then installed at these monitoring points. Anomaly analysis is performed based on the monitoring data from both single-point and zone sensors to determine module anomalies. The system also acquires information on the transmission path nodes of the BIPV system, setting up multiple linkage nodes. Anomaly analysis is performed on the monitoring data from these linkage nodes to determine transmission equipment anomalies. Finally, fire probability prediction is performed based on the module and transmission equipment anomalies, and an early warning is issued based on the prediction results. This combined monitoring by single-point and zone sensors improves the accuracy and efficiency of anomaly detection, thereby enhancing the accuracy of fire early warning systems.

[0065] Example 2

[0066] Based on the same inventive concept as the integrated fire linkage early warning method for building photovoltaics in the foregoing embodiments, such as Figure 2 As shown, this application also provides an integrated fire alarm and early warning system for building photovoltaics, the system comprising:

[0067] The photovoltaic string distribution information acquisition module 11 is used to collect the setting information of N photovoltaic strings of building photovoltaics and obtain the setting planar distribution information.

[0068] The linkage area division module 12 is used to divide multiple photovoltaic group series linkage areas according to the set planar distribution information, and each photovoltaic group series linkage area is equipped with an area monitoring sensor.

[0069] The monitoring point setting module 13 is used to determine the abnormal probability components based on the historical data of the photovoltaic modules of N photovoltaic strings, set the monitoring points of the photovoltaic modules, and set single-point monitoring sensors for the monitoring points of the photovoltaic modules.

[0070] The first anomaly analysis module 14 is used to perform anomaly analysis based on the monitoring data of the single-point monitoring sensor and the area monitoring sensor, and to determine component anomaly information.

[0071] Linkage node setting module 15, the linkage node setting module 15 is used to obtain the transmission path node devices of building photovoltaic and set multiple transmission device linkage nodes;

[0072] The second anomaly analysis module 16 is used to perform anomaly analysis based on the monitoring data of multiple transmission linkage nodes to determine the anomaly information of the transmission equipment.

[0073] Fire probability prediction module 17 is used to predict the probability of fire based on the abnormal information of the components and the abnormal information of the transmission equipment, and to issue an early warning based on the fire probability prediction result.

[0074] Furthermore, the system also includes an associated disconnection control module, which includes:

[0075] Connect to the automatic fire alarm system to obtain fire alarm information;

[0076] Based on the aforementioned planar distribution information, the building coverage rate is calculated, and the buildings are divided into zones based on the building photovoltaic coverage rate.

[0077] The fire range information is determined based on the fire alarm information;

[0078] The correlation probability is determined based on the fire range information, building zoning, and building photovoltaic coverage.

[0079] When the association probability reaches a preset range, association disconnection control information is generated.

[0080] Furthermore, the linkage region division module 12 also includes:

[0081] Acquire the sensing parameters of the area monitoring sensor, including the sensing range and sensing constraint space parameters;

[0082] Based on the aforementioned planar distribution information, the distribution area and the distribution building structure are determined.

[0083] Based on the sensing range and sensing constraint space parameters, interval division constraints are set, and the planar distribution information is divided into regions to obtain multiple photovoltaic group series linkage regions.

[0084] Furthermore, the monitoring point setting module 13 also includes:

[0085] Obtain information on the usage duration, deployment environment, and anomaly records of photovoltaic modules;

[0086] Anomalies are evaluated based on the usage duration, deployment environment, and anomaly records to determine the probability of anomalies.

[0087] Based on the anomaly probability, the anomaly probability component is determined, wherein the anomaly probability component is a photovoltaic module whose anomaly probability reaches a preset probability threshold.

[0088] Furthermore, the first anomaly analysis module 14 also includes:

[0089] The regional monitoring sensor is activated to monitor the interconnected areas of the photovoltaic array and to obtain monitoring data for each interconnected area;

[0090] Based on the positional relationship of each linkage area and the monitoring data of each linkage area, the difference value is calculated, and the anomaly location is determined based on the difference value to identify the abnormal linkage area.

[0091] Based on the abnormal linkage area, the single-point monitoring sensor of the linkage area is located to monitor the photovoltaic module monitoring point and obtain the monitoring data of the photovoltaic module monitoring point.

[0092] Determine whether there are any abnormalities in the monitoring data of the photovoltaic module monitoring points. If so, determine the abnormal information of the module based on the monitoring data of the photovoltaic module monitoring points.

[0093] When no abnormality is found, the abnormality information of the component is determined based on the abnormality linkage area.

[0094] Furthermore, the first anomaly analysis module 14 also includes:

[0095] Start the image acquisition device and acquire images based on the abnormal linkage area to obtain regional image information;

[0096] The image information of the region is segmented according to a preset segmentation region;

[0097] By using a pre-defined abnormal feature convolution kernel to traverse and segment the image, abnormal photovoltaic modules are located, and abnormal information of the modules is determined.

[0098] Furthermore, the system also includes a multi-level response hierarchy analysis module, which includes:

[0099] To obtain the periodic characteristics of building lighting;

[0100] Analysis of the abnormal impact on the series-connected dynamic region of the photovoltaic array based on the aforementioned periodic characteristics;

[0101] Based on the periodic characteristics and the abnormal influence relationship of the photovoltaic array series-interaction area, a multi-level response hierarchy is set for the area sensor and / or single-point monitoring sensor;

[0102] Based on the multi-level response hierarchy, the series-connected dynamic area of ​​the photovoltaic array is monitored in a hierarchical manner;

[0103] When the periodic characteristics change, based on the abnormal influence relationship between the periodic characteristics and the photovoltaic array series-interaction area, the multi-level response hierarchy of the area sensor and / or single-point monitoring sensor is replaced.

[0104] Furthermore, the fire probability prediction module 17 also includes:

[0105] Based on fire accident records, determine the logical impact coefficients between photovoltaic module monitoring data, transmission equipment monitoring data, and fire accidents;

[0106] Based on the logical impact coefficient, the probability of fire is determined by calculating the probability of the component abnormality information and the transmission equipment abnormality information respectively.

[0107] When the fire probability of either component abnormality information or transmission equipment abnormality information reaches a preset threshold, connect to the automatic fire alarm system and send a photovoltaic abnormal fire probability warning.

[0108] Based on the transmission path of building photovoltaics, a transmission impact data chain is constructed, and the probability of fire is used as the impact coefficient of the path nodes.

[0109] Based on the fire accident records, the correlation impact of each transmission path node is obtained, and the correlation impact is added to the impact coefficient as an adjustment coefficient. Probability prediction is performed on each transmission path node, and when the preset threshold is reached, the photovoltaic abnormal fire probability warning is sent.

[0110] The specific example of the integrated fire linkage early warning method for building photovoltaics in the aforementioned embodiment 1 is also applicable to the integrated fire linkage early warning system for building photovoltaics in this embodiment. Through the foregoing detailed description of the integrated fire linkage early warning method for building photovoltaics, those skilled in the art can clearly understand the integrated fire linkage early warning system for building photovoltaics in this embodiment. Therefore, for the sake of brevity, it will not be described in detail here.

[0111] It should be understood that various forms of processes shown above can be used, with steps rearranged, added, or deleted, as long as the desired result of the technical solution disclosed in this application can be achieved, and this document does not impose any restrictions.

[0112] Note that the above are merely preferred embodiments and the technical principles employed in this application. Those skilled in the art will understand that this application is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of this application. Therefore, although this application has been described in detail through the above embodiments, this application is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of this application, the scope of which is determined by the scope of the appended claims.

Claims

1. An integrated fire linkage early warning method for building photovoltaics, characterized in that, include: Collect the configuration information of N photovoltaic strings in building-integrated photovoltaics (BIPV) systems and obtain their planar distribution information. Based on the aforementioned planar distribution information, multiple photovoltaic array series-connected operation areas are divided, and a regional monitoring sensor is installed in each photovoltaic array series-connected operation area. Based on the historical data of photovoltaic modules in N photovoltaic strings, identify modules with abnormal probability, set monitoring points for photovoltaic modules, and set single-point monitoring sensors for the monitoring points of photovoltaic modules. Based on the monitoring data from the single-point monitoring sensor and the area monitoring sensor, anomaly analysis is performed to determine component anomaly information; Obtain the transmission path node devices of building-integrated photovoltaics (BIPV) and set up multiple transmission device linkage nodes; Anomaly analysis is performed based on monitoring data from multiple transmission linkage nodes to determine abnormal information of transmission equipment; Fire probability is predicted based on the abnormal information of the components and transmission equipment, and an early warning is issued based on the fire probability prediction results.

2. The method as described in claim 1, characterized in that, Also includes: Connect to the automatic fire alarm system to obtain fire alarm information; Based on the aforementioned planar distribution information, the building coverage rate is calculated, and the buildings are divided into zones based on the building photovoltaic coverage rate. The fire range information is determined based on the fire alarm information; The correlation probability is determined based on the fire range information, building zoning, and building photovoltaic coverage. When the association probability reaches a preset range, association disconnection control information is generated.

3. The method as described in claim 1, characterized in that, Based on the aforementioned planar distribution information, multiple photovoltaic array series-connected operation zones are defined, and each photovoltaic array series-connected operation zone is equipped with a zone monitoring sensor, including: Acquire the sensing parameters of the area monitoring sensor, including the sensing range and sensing constraint space parameters; Based on the aforementioned planar distribution information, the distribution area and the distribution building structure are determined. Based on the sensing range and sensing constraint space parameters, interval division constraints are set, and the planar distribution information is divided into regions to obtain multiple photovoltaic group series linkage regions.

4. The method as described in claim 1, characterized in that, The method for determining abnormal probability components based on historical data of photovoltaic modules with N photovoltaic strings includes: Obtain information on the usage duration, deployment environment, and anomaly records of photovoltaic modules; Anomalies are evaluated based on the usage duration, deployment environment, and anomaly records to determine the probability of anomalies. Based on the anomaly probability, the anomaly probability component is determined, wherein the anomaly probability component is a photovoltaic module whose anomaly probability reaches a preset probability threshold.

5. The method as described in claim 1, characterized in that, Based on the monitoring data from the single-point monitoring sensor and the area monitoring sensor, anomaly analysis is performed to determine component anomaly information, including: The regional monitoring sensor is activated to monitor the interconnected areas of the photovoltaic array and to obtain monitoring data for each interconnected area; Based on the positional relationship of each linkage area and the monitoring data of each linkage area, the difference value is calculated, and the anomaly location is determined based on the difference value to identify the abnormal linkage area. Based on the abnormal linkage area, the single-point monitoring sensor of the linkage area is located to monitor the photovoltaic module monitoring point and obtain the monitoring data of the photovoltaic module monitoring point. Determine whether there are any abnormalities in the monitoring data of the photovoltaic module monitoring points. If so, determine the abnormal information of the module based on the monitoring data of the photovoltaic module monitoring points. When no abnormality is found, the abnormality information of the component is determined based on the abnormality linkage area.

6. The method as described in claim 5, characterized in that, The step of determining the component's abnormal information based on the abnormal linkage area includes: Start the image acquisition device and acquire images based on the abnormal linkage area to obtain regional image information; The image information of the region is segmented according to a preset segmentation region; By using a pre-defined abnormal feature convolution kernel to traverse and segment the image, abnormal photovoltaic modules are located, and abnormal information of the modules is determined.

7. The method as described in claim 1, characterized in that, Also includes: To obtain the periodic characteristics of building lighting; Analysis of the abnormal impact on the series-connected dynamic region of the photovoltaic array based on the aforementioned periodic characteristics; Based on the periodic characteristics and the abnormal influence relationship of the photovoltaic array series-interaction area, a multi-level response hierarchy is set for the area sensor and / or single-point monitoring sensor; Based on the multi-level response hierarchy, the series-connected dynamic area of ​​the photovoltaic array is monitored in a hierarchical manner; When the periodic characteristics change, based on the abnormal influence relationship between the periodic characteristics and the photovoltaic array series-interaction area, the multi-level response hierarchy of the area sensor and / or single-point monitoring sensor is replaced.

8. The method as described in claim 2, characterized in that, Fire probability prediction is performed based on the abnormal information of the components and transmission equipment, and early warning is issued based on the fire probability prediction results, including: Based on fire accident records, determine the logical impact coefficients between photovoltaic module monitoring data, transmission equipment monitoring data, and fire accidents; Based on the logical impact coefficient, the probability of fire is determined by calculating the probability of the component abnormality information and the transmission equipment abnormality information respectively. When the fire probability of either component abnormality information or transmission equipment abnormality information reaches a preset threshold, connect to the automatic fire alarm system and send a photovoltaic abnormal fire probability warning. Based on the transmission path of building photovoltaics, a transmission impact data chain is constructed, and the probability of fire is used as the impact coefficient of the path nodes. Based on the fire accident records, the correlation impact of each transmission path node is obtained, and the correlation impact is added to the impact coefficient as an adjustment coefficient. Probability prediction is performed on each transmission path node, and when the preset threshold is reached, the photovoltaic abnormal fire probability warning is sent.

9. An integrated fire alarm and early warning system for building photovoltaics, characterized in that, The system is configured to perform the steps of the method according to any one of claims 1 to 8, and comprises: A photovoltaic string distribution information acquisition module is used to collect the setting information of N photovoltaic strings of building photovoltaics and obtain the setting planar distribution information. The linkage area division module is used to divide multiple photovoltaic group series linkage areas according to the set planar distribution information, and each photovoltaic group series linkage area is equipped with an area monitoring sensor. The monitoring point setting module is used to determine the abnormal probability components based on the historical data of the photovoltaic modules of N photovoltaic strings, set the monitoring points of the photovoltaic modules, and set single-point monitoring sensors for the monitoring points of the photovoltaic modules. The first anomaly analysis module is used to perform anomaly analysis based on the monitoring data of the single-point monitoring sensor and the area monitoring sensor to determine component anomaly information. Linkage node setting module, which is used to acquire the transmission path node devices of building photovoltaic and set multiple transmission device linkage nodes; The second anomaly analysis module is used to perform anomaly analysis based on monitoring data from multiple transmission linkage nodes to determine anomaly information of the transmission equipment. A fire probability prediction module is used to predict the probability of fire based on the abnormal information of the components and the abnormal information of the transmission equipment, and to issue an early warning based on the fire probability prediction results.

Images