CN-121997405-A - Pumped storage power station 5G+BIM intelligent interaction management system

CN121997405ACN 121997405 ACN121997405 ACN 121997405ACN-121997405-A

Abstract

The application provides a 5G+BIM intelligent interaction management system of a pumped storage power station, which comprises a sensing layer, a network layer, a data layer and an application layer, wherein the sensing layer is used for collecting earth-rock side vehicle transportation data, dam rolling data, underground cavern construction data and monitoring data, dynamic data updating of a BIM model is realized through a 5G camera, a sensor and construction data, the network layer is used for transmitting the dynamic data of the BIM model through a 5G private network and distributing communication resources according to service requirements by adopting a network slicing strategy, the data layer is used for fusing the dynamic data of the BIM model and comprises BIM geometric data, sensor real-time parameters and construction data, constructing a dynamic digital twin database, the application layer is used for carrying out construction state early warning and construction scheduling with the BIM model based on design parameters and sensor parameters, generating decision instructions through a deep learning algorithm, and realizing data depth fusion and network resource optimization and emergency capability improvement through fusing 5G highly reliable communication, BIM dynamic modeling and AI intelligent decision.

Inventors

DING WANQIN
LIU YANG
WANG WENDONG
SHI ZIHAO
TANG DAOCHU
Xu Motao
LI XIAOHU
SUN YANG
ZHAO PENGYUAN

Assignees

长电(张掖)能源发展有限公司
中国长江电力股份有限公司
中国三峡建工(集团)有限公司
中国电建集团西北勘测设计研究院有限公司

Dates

Publication Date: 20260508
Application Date: 20251203

Claims (10)

1. A pumping energy storage power station 5G+BIM intelligent interaction management system is characterized by comprising: The sensing layer is used for collecting equipment implementation parameters, including real-time parameters of water turbines, generators, water pipelines and reservoir water levels, and realizing dynamic data update of the BIM through a 5G camera, a sensor and underwater inspection equipment; the network layer transmits dynamic data of the BIM model through the 5G private network and adopts a network slicing strategy to allocate communication resources according to service requirements; The data layer is used for fusing the dynamic data of the BIM model and comprises BIM geometric data, equipment real-time parameters and historical operation and maintenance data to construct a dynamic digital twin database; the application layer performs predictive maintenance, intelligent scheduling and risk early warning based on the equipment parameters and the BIM model, and generates a decision instruction through a deep learning algorithm; The edge-cloud cooperative computing module is used for dynamically distributing computing resources according to the real-time requirements of tasks, and performing distributed training and reasoning through an LSTM neural network model and a dispatching model based on reinforcement learning; The edge computing node is deployed in a 5G base station, a field server or a terminal and used for executing tasks with high real-time requirements; The edge computing nodes are divided into a plurality of grades according to the estimated computing power, and different grades correspond to different computing tasks; the cloud computing center provides computing tasks with corresponding orders according to the edge computing nodes and feeds results back to the edge nodes through the 5G private network; the mechanism of the edge-cloud cooperation is specifically as follows: The edge node preprocesses the data and uploads key information to the cloud; the cloud trains the AI model and issues to the edge node to update the local model.
2. The system for intelligently interactively managing the 5G and the BIM of the pumped storage power station according to claim 1 is characterized in that the real-time parameters of the equipment are mapped to a BIM model in real time through a 5G private network to form a dynamic digital twin body; and docking with the dynamic digital twin database by using the BIM API to realize the state update of the real-time parameter driven BIM model.
3. The intelligent interaction management system of the pumped storage power station 5G+BIM according to claim 1 is characterized in that real-time parameters of the water turbine comprise bearing vibration, guide vane opening, water head height and flow; Building a hydraulic turbine fault prediction model, inputting real-time parameters of the hydraulic turbine, predicting the residual life of a bearing by adopting an LSTM neural network model, identifying an abnormal vibration mode by using an abnormal detection algorithm, and outputting potential faults and recommended maintenance schemes 7 days in advance.
4. The system for intelligently managing the 5G and BIM of the pumped storage power station is characterized in that an underwater inspection device is adopted to carry a 5G camera to shoot a video of the inner wall of a pipeline, a crack and cavitation erosion area is detected and the damage area is quantized through a YOLOv8+U-Net combined model, and the detection result is automatically marked to the BIM model to generate a repair work order.
5. The system for intelligently interactively managing the pumped storage power station 5G+BIM according to claim 1 is characterized in that a dispatching model based on reinforcement learning is built, a pumping/power generation switching strategy is optimized according to the power grid load demand and the reservoir water level, input parameters of the dispatching model based on reinforcement learning comprise electricity price, water level and unit efficiency curve, an optimal dispatching instruction is dynamically generated, and the optimal dispatching instruction is issued to a unit control module through a 5G network.
6. The system for intelligently interactively managing the pumped-storage power station 5G+BIM of claim 1 is characterized by further comprising an engineering progress prediction module, which comprises engineering progress prediction and a engineering management function; the project progress prediction function comprises the steps of dynamically adjusting project scheduling according to comprehensive construction progress, predicting construction steps in advance, preparing materials, personnel and equipment, and generating a construction priority work order; The work order management function is specifically to automatically dispatch the work order to the mobile terminal of the responsible person, and remotely guide the installation and maintenance steps by using an AR.
7. The system for intelligently managing the pumped storage power station 5G+BIM according to claim 1 is characterized by further comprising an emergency command module, wherein the emergency command module is used for simulating extreme working conditions based on a BIM model, predicting a dam break influence range through AI and generating an evacuation path; When the backbone network is interrupted, the 5G D2D equipment is started to directly connect to construct a temporary communication network, so that command instruction transmission is ensured.
8. The system for intelligent interaction management of pumped storage power station 5G+BIM according to claim 1, wherein the network slice comprises a control instruction slice, a video monitoring slice and a BIM data slice; The control instruction slice carries a unit start-stop instruction and a valve control instruction; The video monitoring slice carries pictures or video data collected by unmanned aerial vehicle inspection video or underwater inspection equipment; BIM data slices carry the dynamic data of the BIM model and AR/VR interaction data.
9. The pumped storage power station 5g+bim intelligent interaction management system according to claim 1, wherein: Acquiring hardware configuration and work tasks of an edge computing node, dividing the edge computing node into three grades, configuring BIM models of different orders of magnitude in different grades, and setting the work tasks to be the highest priority so as to ensure real-time response; BIM models of different magnitudes refer to a lightweight BIM model and a lightweight local AI model, wherein the current node only comprises a module related to a work task, parameters related to the work task, and a hardware configuration of the current node; the light BIM model is used for reducing redundant data and lower-precision BIM around a work task as much as possible; the lightweight local AI model is only used for completing data preprocessing work related to a work task or only used for executing tasks issued by a cloud computing center without performing iterative training.
10. The pumped storage power station 5G+BIM intelligent interaction management system of claim 1, wherein the distributed training and reasoning process of the LSTM neural network model comprises the following steps: Preprocessing a water turbine bearing vibration signal acquired in real time, extracting key characteristics, and carrying out real-time reasoning by using a lightweight LSTM model to predict the residual life RUL (t) of the bearing, wherein a reasoning formula is as follows: ; Wherein, the In order to hide the state for the LSTM, In order to input the characteristics of the feature, Is an LSTM inference function; Aggregating historical data of a plurality of edge nodes, performing global model training by adopting a federal learning framework, uploading model gradients by the edge nodes, and updating global model parameters by cloud aggregation: ; Wherein, the To represent the global model parameters after cloud aggregation, As an initial parameter of the local model, For the learning rate, N is the number of edge nodes, Model gradients calculated from the local data for the ith edge node; The cloud terminal updates the updated global model parameters And the data are issued to edge nodes to replace local models so as to optimize reasoning accuracy.

Description

Pumped storage power station 5G+BIM intelligent interaction management system Technical Field The invention relates to the field of intelligent interaction management of data, in particular to a 5G+BIM intelligent interaction management system of a pumped storage power station. Background The pumped storage power station is used as an important peak regulation and valley filling facility in a power system, and plays a key role in protecting the stability of a power grid and improving the capacity of renewable energy sources. The construction time of the pumped storage power station is long, the construction procedure is complicated, the construction equipment is numerous, and the existing pumped storage power station management system has obvious defects in the aspects of data integration, decision real-time performance and network adaptability. Moreover, the pumping and accumulating engineering is usually built in mountain areas, the land level difference is large, the terrain is complex, and common public network signals are easily influenced by shielding and reflection, so that the transmission is unstable. On the other hand, the water environment of the pumped storage power station may cause an increase in signal propagation loss, and the network slice may allocate communication resources according to service requirements. If the data collected by the underwater equipment in the pumping and storage engineering is directly transmitted to the cloud, the transmission efficiency may be low due to the attenuation of the water environment signal. Therefore, an integrated management system integrating 5G high-reliability communication, BIM dynamic modeling and AI intelligent decision is needed, which realizes stable signal transmission, deep data integration, intelligent active operation and maintenance, optimizes network resources and improves emergency capability. Disclosure of Invention The invention provides a 5G+BIM intelligent interaction management system of a pumped storage power station for solving the technical problems in the background, which is characterized by comprising a perception layer, a network layer, a data layer and an application layer; The sensing layer is used for collecting real-time parameters of equipment implementation, including water turbine, generator, water pipeline and reservoir water level, and realizing dynamic data update of the BIM model through the 5G camera, the sensor and the underwater inspection equipment; The network layer transmits the dynamic data of the BIM model through the 5G private network, and adopts a network slicing strategy to allocate communication resources according to service requirements; the data layer is used for fusing dynamic data of the BIM model and comprises BIM geometric data, equipment real-time parameters and historical operation and maintenance data to construct a dynamic digital twin database; The application layer performs predictive maintenance, intelligent scheduling and risk early warning based on the equipment parameters and the BIM model, and generates a decision instruction through a deep learning algorithm; the system also comprises an edge-cloud cooperative computing module, a task real-time processing module and a task real-time processing module, wherein the edge-cloud cooperative computing module is used for dynamically distributing computing resources according to the task real-time requirements, and distributed training and reasoning are carried out through an LSTM neural network model and a dispatching model based on reinforcement learning; the edge computing node is deployed in a 5G base station, a field server or a terminal and used for executing tasks with high real-time requirements; the edge computing nodes are divided into a plurality of grades according to the estimated computing power, and different grades correspond to different computing tasks; the cloud computing center is used for executing complex computing tasks and feeding back results to the edge nodes through the 5G private network; The cloud computing center provides computing tasks of corresponding orders according to the edge computing nodes; the mechanism of the edge-cloud cooperation is specifically as follows: The edge node preprocesses the data and uploads key information to the cloud; the cloud trains the AI model and issues to the edge node to update the local model. In the preferred scheme, the real-time parameters of the equipment are mapped to the BIM model in real time through a 5G private network to form a dynamic digital twin body; and docking with the dynamic digital twin database by using the BIM API to realize the state update of the real-time parameter driven BIM model. In a preferred scheme, real-time parameters of the water turbine comprise bearing vibration, guide vane opening, head height and flow; Building a hydraulic turbine fault prediction model, inputting real-time parameters of the hydraulic turbine, predicting the residual life