CN-121982856-A - Mountain photovoltaic falling stone early warning system and method based on multi-mode intelligent monitoring

Abstract

The invention discloses a mountain photovoltaic falling stone early warning system and method based on multi-mode intelligent monitoring, and belongs to the technical field of photovoltaic power station safety monitoring and geological disaster early warning. The system comprises a multi-mode front end monitoring unit, a data transmission network, a central intelligent analysis platform and an intelligent linkage control unit of the photovoltaic power station. The remote monitoring and real-time response of mountain falling rocks are realized by adopting the multi-mode intelligent monitoring, high-precision image recognition and fault diagnosis technology. The invention solves the problems of low rockfall monitoring precision, untimely early warning and lack of linkage with a photovoltaic power station management system in the prior art, and remarkably improves the safety protection capability and the operation and maintenance intelligent level of the mountain photovoltaic power station.

Inventors

• WANG SHUANG
• LIU HAOYU
• PAN DENG
• CHEN YUCHANG
• Zou Xindie
• HU XINGBO
• TAO SHUAI
• LUO HAOPENG
• WU SHANG
• XIAO HUANHUAN
• Xiong Yonger

Assignees

• 三峡集团四川能源投资有限公司
• 中国长江三峡集团有限公司

Dates

Publication Date

20260505

Application Date

20251223

Claims (6)

1. 1. Mountain region photovoltaic falling stone early warning system based on multimode intelligent monitoring, characterized by comprising: The multi-mode front end monitoring unit is used for collecting video, vibration, deformation and environmental data of the slope area; The central intelligent analysis platform is used for carrying out fusion analysis on the data acquired by the multi-mode front end monitoring unit, identifying a falling stone event based on high-precision image identification and multi-source information fault diagnosis technology and generating early warning information; and the intelligent linkage control unit of the photovoltaic power station is used for receiving the early warning information and automatically executing linkage control operation with the photovoltaic power station management system.
2. 2. The system of claim 1, wherein the multi-modal front-end monitoring unit comprises at least two of a high-definition video monitoring module, a three-dimensional laser scanning or millimeter wave radar module, a vibrating fiber optic sensing module, and an environmental sensor module.
3. 3. The system of claim 1 or 2, wherein the central intelligent analysis platform comprises: the data fusion and preprocessing module is used for synchronizing and standardizing the multi-source heterogeneous data; The high-precision image recognition and target tracking module is used for recognizing and tracking falling rocks in the video based on a deep learning algorithm; the multi-source information fault diagnosis and decision module performs comprehensive diagnosis according to a preset rule base by fusing an image recognition result, deformation data, vibration signals and environment data so as to reduce false alarm rate; And the early warning information generation and release module is used for generating hierarchical early warning according to the diagnosis result.
4. 4. The system of claim 3, wherein the rule base of the multi-source information fault diagnosis and decision module comprises determining a high confidence rockfall event when the video identifies a moving object and the vibration sensor detects a matching impact signal within a corresponding spatio-temporal range, while the deformation monitoring data is justified by an object displacement, and the environmental disturbance factor is below a threshold.
5. 5. The system of claim 1, wherein the linkage operation performed by the intelligent linkage control unit of the photovoltaic power plant comprises at least one of: Sending a remote shutdown instruction to power equipment corresponding to the falling stone risk area; controlling a pan-tilt camera to automatically track and record a falling stone event; Triggering an on-site audible and visual alarm device; pushing the early warning information and the treatment record to a power station operation and maintenance management platform.
6. 6. The mountain land photovoltaic rockfall early warning method based on multi-mode intelligent monitoring is characterized by comprising the following steps of: Acquiring slope multisource monitoring data in real time through a multi-mode front end monitoring unit; The data are fused and intelligently analyzed through a central intelligent analysis platform, so that the falling stone event is accurately diagnosed; And generating early warning information, and triggering automatic linkage response with a power station management system through the intelligent linkage control unit of the photovoltaic power station.

Description

Mountain photovoltaic falling stone early warning system and method based on multi-mode intelligent monitoring Technical Field The invention relates to the technical field of safety monitoring and geological disaster early warning of photovoltaic power stations, in particular to a falling stone early warning system and method suitable for mountain environments. Background With the advancement of the goal of carbon neutralization, the photovoltaic power generation industry has evolved rapidly. The mountain photovoltaic is widely applied because of the advantages of no occupation of cultivated land, high cleaning efficiency and the like. However, mountain terrain is complex, geological disasters occur frequently, and especially slope falling rocks form a serious threat to the safety of photovoltaic panels, support foundations, cables and operation and maintenance personnel. The existing falling rock monitoring technology mainly has the following defects: The monitoring precision is low, the false alarm rate is high, a single sensor (such as a vibration sensor) is easy to be interfered by environmental noise (such as wind and rain, animal activities and equipment operation vibration), and the falling rock event is difficult to accurately identify, so that the early warning reliability is poor. The early warning is not timely, the traditional monitoring means (such as manual inspection and simple sensors) are slow in response, long delay exists from the discovery of the abnormality to the sending of the early warning, and enough time cannot be reserved for emergency response. The intelligent monitoring system is lack of effective linkage with a photovoltaic power station management system, wherein most of existing monitoring systems are information islands, and early warning information cannot directly and automatically trigger linkage actions of core control systems (such as box transformer, inverter and security system) of the power station. For example, the power supply of the corresponding inverter cannot be automatically cut off in a high-risk area of falling rocks to prevent electric secondary accidents, or the camera cannot be linked to perform automatic tracking video recording, so that the emergency treatment efficiency is low, and the safety risk is increased. Therefore, there is an urgent need for a falling rock early warning solution that can achieve high precision, real-time early warning, and can be deeply integrated with a photovoltaic power station management system. Disclosure of Invention The invention aims to overcome the defects of the prior art, provides a mountain photovoltaic falling stone early warning system and a mountain photovoltaic falling stone early warning method, the system adopts a multi-mode intelligent monitoring and high-precision image recognition technology, the accuracy and timeliness of falling rock identification are remarkably improved, and automatic closed-loop management from early warning to treatment is realized through deep linkage with a photovoltaic power station management system. In order to achieve the above purpose, the invention adopts the following technical scheme: Mountain region photovoltaic falling stone early warning system, characterized in that includes: 1. The multi-mode front end monitoring unit is deployed at a key position of a photovoltaic field side slope and comprises: (1) And the high-definition video monitoring module is used for collecting real-time video image data of the slope area. (2) The three-dimensional laser scanning module/millimeter wave radar module is used for carrying out high-frequency and high-precision three-dimensional deformation monitoring on the surface of the side slope and perceiving tiny displacement and vibration. (3) And the vibration optical fiber sensing module is laid along the photovoltaic bracket or the cable trench and is used for sensing vibration signals generated by falling stone impact and rolling. (4) The environment sensor module is used for collecting environment data such as wind speed, rainfall and the like and providing environment background information for subsequent data fusion and fault diagnosis. 2. And the data transmission network adopts a mode of combining wired (such as optical fiber) and wireless (such as 4G/5G, loRa) to transmit the multi-mode data acquired by the front-end monitoring unit to the central processing platform in real time and reliably. 3. Central intelligent analysis platform, as the core of the system, comprising: (1) And the data fusion and preprocessing module is used for carrying out time synchronization, filtering and standardization processing on the received multi-source heterogeneous data such as video, radar, vibration, environment and the like. (2) And the high-precision image recognition and target tracking module is used for analyzing the video stream in real time, precisely recognizing and tracking the falling stone target in motion and estimating the motion tr