Intelligent inspection method and system for hydropower station

Abstract

The invention relates to the technical field of hydropower station inspection, in particular to an intelligent hydropower station inspection method and system, and the method comprises the steps: enabling an inspection terminal to be connected with to-be-inspected equipment through a short-distance communication protocol, and enabling the intelligent inspection terminal to collect the attribute information of the to-be-inspected equipment and the personal data of a to-be-assigned inspector; the inspection terminal retrieves all tasks related to the to-be-inspected equipment in the current inspection time range from preset task data, and the tasks currently executable by the to-be-assigned inspector are screened from the tasks to form a first execution scheme; the inspection terminal uses a cloud rendering engine to construct an optimization algorithm three-dimensional model for the operation state data and the historical data, and a plurality of alternative execution schemes are generated through simulation combination of the optimization algorithm three-dimensional model; and comparing the first execution scheme with the plurality of alternative execution schemes, and obtaining the alternative execution scheme with the highest overlap ratio as a second execution scheme. According to the invention, errors and time consumption caused by manual recording can be avoided, and inspection efficiency and data accuracy are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of hydropower station inspection, and more specifically, to an intelligent inspection method and system for hydropower stations. Background Art

[0002] Improving the intelligent level of hydropower stations has become an inevitable trend in the industry development, and the intelligent upgrade of equipment inspection work is particularly crucial. The traditional manual inspection method not only requires a large amount of human and time costs, but is also extremely vulnerable to the subjective judgment of inspectors, resulting in situations such as missed reports, false reports, and misstatements. These factors may all lead to deviations in inspection results and even omission of key potential safety hazards. Therefore, the efficiency and accuracy of traditional manual inspection are difficult to meet the increasingly strict safety supervision requirements of hydropower stations, and it is also difficult to adapt to the current situation of the continuous expansion of hydropower station scale and the increasing complexity of technology. Summary of the Invention

[0003] The purpose of the present invention is to provide an intelligent inspection method and system for hydropower stations, which quickly connect the inspection terminal with the equipment to be inspected through short-distance communication, automatically collect the configuration parameters, operating status, historical data, and location information of the equipment, avoiding the errors and time consumption of manual recording, and improving the inspection efficiency and data accuracy.

[0004] The present invention is realized through the following technical solutions:

[0005] An intelligent inspection method for hydropower stations, the steps of the method include:

[0006] Establish a connection between the inspection terminal and the equipment to be inspected through a short-distance communication protocol. The intelligent inspection terminal collects the attribute information of the equipment to be inspected and the personal data of the inspector to be assigned. The attribute information includes the configuration parameters, operating status data, historical data, and location data of the equipment to be inspected, and the personal data includes the location data and moving speed of the inspector to be assigned;

[0007] The inspection terminal retrieves all tasks within the current inspection time range and related to the equipment to be inspected from the preset task data, and screens out the tasks that the inspector to be assigned can currently execute to form a first execution plan;

[0008] The inspection terminal uses a cloud rendering engine to construct an optimized algorithm three-dimensional model with the operating status data and historical data, obtains a first data set through simulation combination of the optimized algorithm three-dimensional model, and generates multiple alternative execution plans based on the first data set and the location data of the equipment to be inspected;

[0009] Compare the first execution plan with multiple alternative execution plans, and obtain the alternative execution plan with the highest coincidence degree as the second execution plan.

[0010] Preferably, the implementation solution includes an inspection route corresponding to the device to be inspected, and the method for determining the inspection route is as follows:

[0011] Obtain route planning data, where the route planning data includes: the number N of inspectors to be assigned and the initial position data (x i , y i ) of each inspector to be assigned, i = 1, 2... N, the number M of devices to be inspected, and the geographical location of the devices to be inspected is (x j , y j ), j = 1, 2... M, the inspection duration t j , and the device priority p j ;

[0012] Dynamically allocate cloud computing resources according to the optimized algorithm three-dimensional model, and render the inspection route planning data in the cloud according to the allocated cloud computing resources.

[0013] Preferably, the optimized algorithm three-dimensional model is as follows:

[0014]

[0015] Among them, d(r i , r i+1 ) represents the path distance between two adjacent devices to be inspected, (i = 1, 2... N), v is the moving speed of the inspector to be assigned, (j = 1, 2... M), α is the priority weight coefficient of the device to be inspected, Ε is the inspection route, and Ε takes the minimum value.

[0016] Preferably, when rendering the inspection route planning data in the cloud according to the allocated cloud computing resources, the following steps are specifically included:

[0017] Based on real-time data stream processing technology, perform real-time optimization and adjustment on the historical data;

[0018] Design a dynamic model update mechanism based on a deep learning model and an enhanced learning algorithm, continuously optimize the digital twin model using historical data, and correct potential deviations through comparison and calibration with the operating state data.

[0019] Preferably, when designing the dynamic model update mechanism based on a deep learning model and an enhanced learning algorithm, the following steps are specifically included:

[0020] Analyze the complexity and data volume of the digital twin three-dimensional model to obtain the required resource type and quantity;

[0021] Obtain the available quantity and available type in the current cloud computing resources, and match them with the required quantity and the required resource type respectively;

[0022] Classify multiple data centers using the K-means clustering algorithm to obtain the optimal data center and multiple sub-optimal data centers;

[0023] Use the optimal data center to read the rendering task queue and prioritize multiple tasks in the rendering task queue to obtain multiple task priorities, where the task priorities correspond to the tasks one by one;

[0024] Obtain the network latency time corresponding to each task, and adjust the task priorities according to all the network latency times to obtain multiple latest task priorities, so that the tasks corresponding to shorter network latency times have higher latest task priorities;

[0025] Allocate all the tasks to the data centers in sequence according to the multiple latest task priorities, with one task allocated to one data center.

[0026] Preferably, the data center is a collection of physical servers and network infrastructure on a cloud computing platform, and the sub-optimal data centers are the data centers other than the optimal data center.

[0027] An intelligent inspection system for hydropower stations, comprising:

[0028] A real-time acquisition and analysis module for obtaining the attribute information of the equipment to be inspected and the personal data of the inspectors to be assigned and performing data compatibility processing;

[0029] A digital twin configuration module for constructing an optimized algorithm three-dimensional model based on the attribute information of the equipment to be inspected and the personal data of the inspectors to be assigned;

[0030] An allocation module for dynamically allocating cloud computing resources according to the digital twin three-dimensional model and performing rendering in the cloud according to the allocated cloud computing resources;

[0031] A visualization management module for transmitting the data obtained by the acquisition unit to the host computer and performing interactive visual display through the host computer.

[0032] Preferably, the real-time acquisition and analysis module further includes an industrial network protocol compatibility unit, a control system data compatibility unit, a communication service data compatibility unit, a relational database integration support unit, and a second data processing unit;

[0033] The industrial network protocol compatibility unit is used for interface adaptation and data acquisition of multiple industrial network protocols to obtain the operation parameter data of the equipment to be inspected in real time;

[0034] The control system data compatibility unit is used to connect and interact with the DCS control system, PLC control system, and auxiliary network control system of the equipment to be inspected, so as to obtain the control status, monitoring signals, and operation records of the equipment to be inspected;

[0035] The communication service data compatibility unit is used to support HTTP communication protocol, WebSocket communication protocol, and gRPC communication protocol to obtain the historical data and configuration parameters of the equipment to be inspected;

[0036] The relational database integration support unit is used to perform integrated interaction on multiple relational databases to realize the storage, management, and query of the operation data of the hydropower station;

[0037] The second data processing unit is used to process the data obtained by the industrial network protocol compatibility unit, control system data compatibility unit, communication service data compatibility unit, or relational database integration support unit to achieve the compatibility of different types of data.

[0038] The technical solution of the present invention has at least the following advantages and beneficial effects:

[0039] The present invention quickly connects the inspection terminal and the equipment to be inspected through short-distance communication, automatically collects the configuration parameters, operation status, historical data, and location information of the equipment, avoiding the errors and time-consuming of manual records, and improving the inspection efficiency and data accuracy. Secondly, combining the real-time location and speed information of the inspector, automatically screening out the tasks that the inspector can execute, and based on the optimized algorithm three-dimensional model constructed by the cloud rendering engine, simulating multiple inspection routes, providing multiple alternative plans, and finally screening out the plan with the highest coincidence degree with the optimal plan, ensuring the rationality and efficiency of the inspection tasks. This can not only shorten the inspection time, reduce labor costs, but also improve the timeliness and accuracy of equipment maintenance, avoid equipment failures caused by insufficient inspections, ensure the safe and stable operation of the hydropower station, and meet the increasingly strict safety supervision requirements of the hydropower station. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a schematic flowchart of a hydropower station intelligent inspection method provided by the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0041] In order to make the purpose, technical solution, and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below in conjunction with the drawings in the present invention. Obviously, what is described is a part of the present invention, not all of it. Usually, the components of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0042] Such as Figure 1As shown, the present invention provides one of the embodiments: an intelligent inspection method for a hydropower station, and the steps of the method include:

[0043] Through a short - distance communication protocol, establish a connection between the inspection terminal and the equipment to be inspected. The intelligent inspection terminal collects the attribute information of the equipment to be inspected and the personal data of the inspector to be assigned. The attribute information includes the configuration parameters, operation status data, historical data, and location data of the equipment to be inspected, and the personal data includes the location data and moving speed of the inspector to be assigned;

[0044] The inspection terminal retrieves all tasks within the current inspection time range and related to the equipment to be inspected from the preset task data, and screens out the tasks that the inspector to be assigned can currently execute to form a first execution plan;

[0045] The inspection terminal uses a cloud rendering engine to construct an optimized algorithm 3D model with the operation status data and historical data, obtains a first data set through simulation combination of the optimized algorithm 3D model, and generates multiple alternative execution plans based on the first data set and the location data of the equipment to be inspected;

[0046] Compare the first execution plan with multiple alternative execution plans, and obtain the alternative execution plan with the highest coincidence degree as the second execution plan.

[0047] In this embodiment, the inspection terminal is quickly connected to the equipment to be inspected through short - distance communication, automatically collecting the configuration parameters, operation status, historical data, and location information of the equipment, avoiding the errors and time - consuming of manual recording, and improving the inspection efficiency and data accuracy. Secondly, combining the real - time location and speed information of the inspector, automatically screening out the tasks that the inspector can execute, and based on the optimized algorithm 3D model constructed by the cloud rendering engine, simulating multiple inspection routes, providing multiple alternative plans, and finally screening out the plan with the highest coincidence degree with the optimal plan, ensuring the rationality and efficiency of the inspection task. This can not only shorten the inspection time, reduce labor costs, but also improve the timeliness and accuracy of equipment maintenance, avoid equipment failures caused by insufficient inspection, ensure the safe and stable operation of the hydropower station, and meet the increasingly strict safety supervision requirements of the hydropower station.

[0048] More specifically, the execution plan includes the inspection route corresponding to the equipment to be inspected, and the method for determining the inspection route is:

[0049] Obtain route planning data, where the route planning data includes: the number N of inspectors to be assigned and the initial location data (x i , y i ) of each inspector to be assigned, i = 1, 2... N, the number M of equipment to be inspected, and the geographical location of the equipment to be inspected is (x j , y j), j = 1, 2…M, inspection duration t j , equipment priority p j ;

[0050] Dynamically allocate cloud computing resources according to the optimized algorithm three-dimensional model, and render the inspection route planning data in the cloud according to the allocated cloud computing resources.

[0051] The optimized algorithm three-dimensional model is as follows:

[0052]

[0053] Among them, d(r i , r i+1 ) represents the path distance between two adjacent devices to be inspected, (i = 1, 2...N), v is the moving speed of the inspector to be assigned, (j = 1, 2...M), α is the priority weight coefficient of the device to be inspected, Ε is the inspection route, and Ε takes the minimum value.

[0054] In the specific application of this embodiment, when rendering the inspection route planning data in the cloud according to the allocated cloud computing resources, the following steps are specifically included:

[0055] Based on real-time data stream processing technology, perform real-time optimization and adjustment on the historical data;

[0056] Design a dynamic model update mechanism based on deep learning models and reinforcement learning algorithms, continuously optimize the digital twin model using historical data, and correct potential deviations through comparison and calibration with operation status data.

[0057] By combining cloud rendering technology with real-time data stream processing, deep learning, and reinforcement learning algorithms, the intelligence and accuracy of inspections can be significantly improved. Additionally, by combining real-time optimization and adjustment of historical data, the digital twin model can more accurately reflect the current state of the equipment and predict potential problems. At the same time, the introduction of deep learning models enables the system to learn from historical data, continuously optimize the digital twin model, and correct deviations through comparison and calibration with real-time operation data, ensuring the authenticity of the simulation and the accuracy of predictions. This not only helps inspectors understand the equipment status more comprehensively and deeply but also assists them in predicting potential failures, thereby enabling more precise and timely maintenance of the equipment, effectively reducing the failure rate, improving the operation efficiency and safety of the hydropower station, and simultaneously reducing unnecessary maintenance costs.

[0058] In this embodiment, when designing the dynamic model update mechanism for deep learning models and reinforcement learning algorithms, the following steps are specifically included:

[0059] Analyze the complexity and data volume of the digital twin three-dimensional model to obtain the required resource types and quantities;

[0060] Obtain the available quantity and available type in the current cloud computing resources, and respectively match them with the required quantity and the required resource type;

[0061] Use the K-means clustering algorithm to classify multiple data centers to obtain the optimal data center and multiple sub-optimal data centers;

[0062] Use the optimal data center to read the rendering task queue and prioritize multiple tasks in the rendering task queue to obtain multiple task priorities, where the task priorities correspond to the tasks one by one;

[0063] Obtain the network latency time corresponding to each task, and adjust the task priorities according to all the network latency times to obtain multiple updated task priorities, so that the tasks corresponding to shorter network latency times have higher updated task priorities;

[0064] Allocate all the tasks to the data centers in sequence according to the multiple updated task priorities, and allocate one task to one data center.

[0065] More specifically, the data center is a collection of physical servers and network infrastructure on a cloud computing platform, and the sub-optimal data center is the data center other than the optimal data center.

[0066] The above dynamic model update mechanism is of great significance for improving the inspection efficiency and resource utilization rate. First, by analyzing the complexity and data volume of the digital twin three-dimensional model, the demand for cloud computing resources is accurately matched, avoiding resource waste or shortage. Secondly, the K-means clustering algorithm is used to classify the data centers and select the optimal data center to ensure the efficiency of the rendering tasks. More importantly, by dynamically adjusting the task priorities according to the network latency time and giving priority to tasks with lower latency, the rendering speed and the overall inspection response speed are further improved. Finally, the tasks are allocated to each data center to achieve parallel processing, accelerating the update speed of the digital twin model. This refined resource management and task scheduling can make the hydropower station inspection system more efficient, timely reflect the equipment status, assist the inspection personnel to make more reasonable decisions, thereby effectively improving the equipment maintenance efficiency, reducing the failure risk, and ultimately ensuring the stable operation of the hydropower station.

[0067] Furthermore, the present invention also provides a hydropower station intelligent inspection system, which includes the following:

[0068] The real-time acquisition and analysis module is used to obtain the attribute information of the device to be inspected and the personal data of the inspector to be assigned, and perform data compatibility processing;

[0069] The digital twin configuration module is used to construct an optimized algorithm 3D model based on the attribute information of the device to be inspected and the personal data of the inspector to be assigned;

[0070] The allocation module is used to dynamically allocate cloud computing resources according to the digital twin 3D model, and perform rendering in the cloud according to the allocated cloud computing resources;

[0071] The visualization management module is used to transmit the data obtained by the acquisition unit to the host computer and perform interactive visual display through the host computer.

[0072] More specifically, the real-time acquisition and analysis module further includes an industrial network protocol compatibility unit, a control system data compatibility unit, a communication service data compatibility unit, a relational database integration support unit, and a second data processing unit;

[0073] The industrial network protocol compatibility unit is used to perform interface adaptation and data acquisition on multiple industrial network protocols to obtain the operation parameter data of the device to be inspected in real time;

[0074] The control system data compatibility unit is used to connect and interact with the DCS control system, PLC control system, and auxiliary network control system of the device to be inspected to obtain the control status, monitoring signals, and operation records of the device to be inspected;

[0075] The communication service data compatibility unit is used to support HTTP communication protocol, WebSocket communication protocol, and gRPC communication protocol to obtain the historical data and configuration parameters of the device to be inspected;

[0076] The relational database integration support unit is used to perform integrated interaction on multiple relational databases to realize the storage, management, and query of the operation data of the hydropower station;

[0077] The second data processing unit is used to process the data obtained by the industrial network protocol compatibility unit, the control system data compatibility unit, the communication service data compatibility unit, or the relational database integration support unit to achieve the compatibility of different types of data.

[0078] The above is only the preference of the present invention and is not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A hydropower station intelligent inspection method, characterized in that: The steps of the method include: Through a short-distance communication protocol, the inspection terminal is connected to the device to be inspected, and the intelligent inspection terminal collects attribute information of the device to be inspected and personal data of the inspector to be assigned, wherein the attribute information includes configuration parameters, operation status data, historical data and location data of the device to be inspected, and the personal data includes location data and moving speed of the inspector to be assigned; The inspection terminal retrieves all tasks related to the equipment to be inspected within the inspection time range from the preset task data, selects the tasks that can be currently executed by the inspector to be assigned, and forms a first execution plan; The inspection terminal uses a cloud rendering engine to construct an optimization algorithm three-dimensional model with the operating status data and historical data, obtains a first data set after simulation and combination of the optimization algorithm three-dimensional model, and generates multiple alternative execution plans based on the first data set and the location data of the device to be inspected; The first execution scheme is compared with a plurality of alternative execution schemes, and the alternative execution scheme with the highest degree of overlap is obtained as the second execution scheme.

2. The intelligent inspection method for a hydropower station according to claim 1, characterized in that: The execution scheme includes an inspection route corresponding to the equipment to be inspected, and the method for determining the inspection route is: Obtain route planning data, the route planning data including: the number of inspectors to be assigned N and the initial position data (x i ,y i ), i = 1, 2…N, the number of devices to be inspected is M, and the geographical location of the devices to be inspected is (x j ,y j ), j = 1, 2…M, inspection duration t j , device priority p j ; Cloud computing resources are dynamically allocated according to the optimization algorithm three-dimensional model, and the inspection route planning data is rendered in the cloud according to the allocated cloud computing resources.

3. The intelligent inspection method for a hydropower station according to claim 2, characterized in that: The three-dimensional model of the optimization algorithm is as follows: Among them, d(r i ,r i+1 ) represents the path distance between two adjacent devices to be inspected, (i=1,2...N), v is the moving speed of the inspector to be assigned, (j=1,2...M), α is the priority weight coefficient of the device to be inspected, E is the inspection route, and E takes the minimum value.

4. The intelligent inspection method for a hydropower station according to claim 2, characterized in that: When the inspection route planning data is rendered in the cloud according to the allocated cloud computing resources, the following steps are specifically included: Based on real-time data stream processing technology, the historical data is optimized and adjusted in real time; A dynamic model update mechanism is designed based on deep learning models and reinforcement learning algorithms, and historical data is used to continuously optimize the digital twin model. Potential deviations are corrected through comparison and calibration with operating status data.

5. The intelligent inspection method for a hydropower station according to claim 4, characterized in that: When designing a dynamic model update mechanism for a deep learning model and a reinforcement learning algorithm, the following steps are specifically included: Analyze the complexity and data volume of the digital twin three-dimensional model to obtain the required resource type and required quantity; Obtaining the available amount and available type of current cloud computing resources, and matching them with the required amount and the required resource type respectively; The K-means clustering algorithm is used to classify multiple data centers to obtain the optimal data center and multiple suboptimal data centers; Using the optimal data center to read a rendering task queue and prioritize multiple tasks in the rendering task queue to obtain multiple task priorities, where the task priorities correspond to the tasks one by one; Obtaining the network delay time corresponding to each of the tasks, adjusting the task priority according to all the network delay times, and obtaining a plurality of latest task priorities, so that the latest task priority corresponding to the task with a shorter network delay time is higher; All the tasks are sequentially allocated to the data centers according to the multiple latest task priorities, with one task being allocated to one data center.

6. The intelligent inspection method for a hydropower station according to claim 5, characterized in that: The data center is a collection of physical servers and network infrastructure on a cloud computing platform, and the suboptimal data center is the data center other than the optimal data center.

7. An intelligent inspection system for a hydropower station, characterized in that: include: Real-time collection and analysis module, used to obtain the attribute information of the equipment to be inspected and the personal data of the inspectors to be assigned and perform data compatibility processing; A digital twin configuration module, used to construct an optimization algorithm three-dimensional model based on the attribute information of the equipment to be inspected and the personal data of the inspector to be assigned; An allocation module, used to dynamically allocate cloud computing resources according to the digital twin three-dimensional model, and render in the cloud according to the allocated cloud computing resources; The visualization management module is used to transmit the data acquired by the acquisition unit to a host computer and perform interactive visual display through the host computer.

8. The intelligent inspection system for a hydropower station according to claim 7 is characterized in that: The real-time acquisition and analysis module also includes an industrial network protocol compatibility unit, a control system data compatibility unit, a communication service data compatibility unit, a relational database integration support unit and a second data processing unit; The industrial network protocol compatibility unit is used to perform interface adaptation and data collection on a variety of industrial network protocols to obtain operating parameter data of the equipment to be inspected in real time; The control system data compatibility unit is used for connecting and exchanging data between the DCS control system, PLC control system and auxiliary network control system of the equipment to be inspected, so as to obtain the control status, monitoring signal and operation record of the equipment to be inspected; The communication service data compatibility unit is used to support HTTP communication protocol, WebSocket communication protocol and gRPC communication protocol to obtain historical data and configuration parameters of the device to be inspected; The relational database integration support unit is used to integrate and interact with multiple relational databases to achieve storage, management and query of the operation data of the hydropower station; The second data processing unit is used to process the data obtained by the industrial network protocol compatibility unit, the control system data compatibility unit, the communication service data compatibility unit or the relational database integration support unit to achieve compatibility of different types of data.

Images