CN-122026802-A - AI tracker intelligent platform

CN122026802ACN 122026802 ACN122026802 ACN 122026802ACN-122026802-A

Abstract

The invention relates to the field of intelligent operation and maintenance of photovoltaics, and particularly discloses an intelligent platform of an AI tracker, which comprises a data acquisition module, a data analysis module, a control strategy generation module and a remote control execution module which are sequentially connected, wherein the data acquisition module acquires operation and environment data from an NCU/TCU unit of the tracker, the data analysis module performs fault diagnosis and health assessment through an AI model, the control strategy generation module generates a control instruction according to a health report and a time-of-use electricity price, and the remote control execution module completes instruction issuing and feedback closed loop. The intelligent photovoltaic tracker system and the method support private local deployment, public cloud deployment and local and public cloud mixed deployment of the power station, integrate Beidou communication, have an active protection strategy based on prediction and a nationwide software system, solve the problems of large control delay, passive operation and maintenance mode, insufficient safety protection, inflexible deployment and the like of the existing system, and realize intelligent, high-reliability and income maximization operation of the photovoltaic tracker.

Inventors

WANG GUANGMING
XU LE
LI SHANGDONG
DING FENGJUN
ZHU JIANJING

Assignees

杭州帷盛科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260121

Claims (10)

1. An AI tracker intelligent platform, comprising: The data acquisition module is used for acquiring operation data and environment data from an NCU or TCU unit of the photovoltaic tracker; the data analysis module is used for receiving the operation data and the environment data, performing fault diagnosis and state evaluation through an AI model and generating a health evaluation report; the control strategy generation module is used for generating a tracker control instruction according to the health evaluation report and the time-of-use electricity price data; and the remote control execution module is used for sending the tracker control instruction to the corresponding NCU or TCU unit to realize the adjustment of the tracker angle.
2. The intelligent AI tracker platform of claim 1, wherein the data acquisition module specifically comprises: The equipment data acquisition unit is used for acquiring running state data and electrical parameters including tracker angles, motor current and voltage, driving states and fault codes in real time through sensors built in the NCU or the TCU unit, and carrying out primary structural packaging on the acquired data; the environment data acquisition unit is used for connecting a meteorological station deployed by a photovoltaic power station or accessing an authoritative meteorological data service interface, and acquiring environment data and geographic information data comprising irradiance, environment temperature, wind speed and direction, rainfall and snowvolume and component temperature; The protocol adaptation unit is used for dynamically selecting or parallelly enabling communication protocols according to a power station deployment mode and network conditions, wherein the Zigbee protocol or the Lora protocol is preferably enabled under a local area network environment, the 4G or 5G mobile network protocol is enabled when remote public network communication is required, and the Beidou short message communication protocol is enabled under the condition of no public network signal or strict requirements on communication autonomy.
3. The intelligent platform of AI tracker of claim 1, wherein the data analysis module specifically comprises: The data preprocessing unit is used for cleaning, normalizing and extracting features of the operation data and the environment data from the data acquisition module to form a structured analysis data set; The AI fault diagnosis unit is used for receiving the structured analysis data set and identifying potential fault types and abnormal operation modes of the NCU or the TCU unit through a pre-trained AI analysis model; The state evaluation unit is used for evaluating the real-time health state and performance decline trend of the equipment based on the potential fault type and the abnormal operation mode and combining the historical operation data of the equipment; and the report generation unit is used for summarizing the potential fault types, the abnormal operation modes, the real-time health states and the performance decline trend, generating a health evaluation report comprising a fault early warning list, a health state overview and a maintenance priority suggestion, and outputting the health evaluation report to the control strategy generation module.
4. The intelligent platform of an AI tracker of claim 3, wherein the workflow of the AI fault diagnosis unit comprises: receiving a structured analysis data set from the data preprocessing unit, and calling a pre-trained machine learning model, wherein the machine learning model is trained through historical fault data, and integrates a device operation mechanism and a data driving mode; The structured analysis data set is input into a machine learning model, the machine learning model firstly carries out sliding window feature extraction on time sequence data to obtain time sequence feature vectors, then carries out pattern matching and probability calculation based on the time sequence feature vectors, outputs confidence probability of each type of preset faults, and marks the faults with the probability exceeding a set threshold as potential fault types; For operational data that does not directly match the preset fault type but the feature vector deviates significantly from the normal baseline, the abnormal operational mode is marked.
5. The intelligent platform of AI tracker of claim 4, wherein the workflow of the state evaluation unit comprises: Receiving potential fault types, corresponding confidence probabilities and abnormal operation modes from the AI fault diagnosis unit, inquiring a preset equipment influence weight library according to the potential fault types and the abnormal operation modes, and determining a quantitative influence coefficient of each fault or abnormality on the overall health of the equipment; calling a historical performance baseline of the equipment, comparing the current operation data with corresponding performance data under the same working condition in the same period of the history, and calculating a performance deviation rate; Based on the quantized influence coefficient, the confidence probability and the performance deviation rate, inputting the quantized influence coefficient, the confidence probability and the performance deviation rate into a decay model based on a health index for calculation, outputting quantized real-time health state scores, and analyzing trends according to time sequence data to generate a performance decay trend prediction curve.
6. The intelligent AI tracker platform of claim 1, wherein the control strategy generation module specifically comprises: The strategy input unit is used for receiving the health evaluation report, synchronously accessing the time-of-use electricity price data stream issued by the electricity market, and carrying out time stamp alignment and formatting encapsulation on the time-of-use electricity price data stream and the time-of-use electricity price data stream; the active protection strategy unit is used for generating an active risk avoiding instruction comprising a strong wind protection strategy and a strong snow protection strategy according to real-time or predicted meteorological data; The power generation gain optimization unit is used for fusing the equipment state information and the time-of-use electricity price data in the health evaluation report, calculating through a gain maximization model and generating a tracker angle adjustment instruction oriented to power generation gain optimization; And the instruction fusion and issuing unit is used for carrying out priority arbitration and conflict resolution on the active risk avoidance instruction and the angle adjustment instruction, generating a final tracker control instruction and outputting the final tracker control instruction to the remote control execution module.
7. The intelligent platform of AI tracker of claim 6, wherein the workflow of the active protection policy unit comprises: Continuously receiving formatted meteorological data from the strategy input unit and short-term forecast meteorological data from a meteorological server; Comparing the real-time wind speed data with a big wind protection threshold value, and simultaneously comparing the real-time snowfall data or the predicted snowfall data with the big snow protection threshold value, and judging that active protection needs to be triggered when the real-time wind speed exceeds the big wind protection threshold value or the real-time snowfall and the predicted snowfall exceed the big snow protection threshold value; According to the triggered protection type, a corresponding preset safety angle is called from a preset safety strategy library, and the safety angle is calculated and determined based on the structural strength of the tracker and a local wind pressure and snow load model; and monitoring meteorological conditions in real time until the wind speed or snowfall data fall below a safety threshold value and maintain the preset time, generating an instruction for removing the protection state, and returning control rights to the power generation income optimization unit.
8. The AI tracker intelligent platform of claim 7, wherein the workflow of the generation revenue optimization unit comprises: Continuously monitoring whether a forced execution instruction is triggered, and if not, receiving a health assessment report and real-time and predicted time-of-use electricity price data from a strategy input unit; The allowable tracking angle range, the maximum angular speed limit, the real-time irradiance data and the time-of-use electricity price data are used as boundary conditions and input parameters and are input into a benefit maximization model; The profit maximization model aims at maximization of expected power generation in a future scheduling period, rolling optimization calculation is carried out based on the photovoltaic power generation physical model and the electricity price time sequence, and an optimal tracker angle time sequence is solved; and converting the optimal tracker angle time sequence into a specific angle adjustment instruction with a time stamp, and outputting the angle adjustment instruction to an instruction fusion and issuing unit.
9. The intelligent AI tracker platform of claim 1, wherein the remote control execution module specifically comprises: The instruction receiving and checking unit is used for receiving the tracker control instruction from the instruction fusion and issuing unit and checking the format validity and the validity of the target equipment address; the communication protocol adaptation and encapsulation unit is used for encapsulating the tracker control instruction passing the verification into a data frame of a corresponding protocol according to the network type and the communication protocol accessed by the target NCU or the TCU unit; The instruction issuing unit is used for sending the data frame to the target NCU or the TCU unit through a corresponding physical communication link, and driving the tracker to execute an angle adjustment action; The execution feedback acquisition unit is used for receiving the execution state feedback data from the NCU or the TCU unit in real time after the instruction is issued, and returning the execution state feedback data to the data analysis module for updating the equipment state and forming a control closed loop.
10. The intelligent platform of claim 9, wherein the workflow for executing the feedback collection unit specifically comprises: Immediately starting a feedback waiting timer for the current instruction after the instruction issuing unit successfully sends a data frame, and marking the instruction as an executing state; Matching and checking the received feedback data with the issued original instruction, and judging that the instruction is successfully executed if the deviation between the actual execution angle and the target angle is within the allowable error range and the status code is successful; If feedback is not received before the feedback waiting timer is overtime, or the feedback status code is failed, or the actual execution angle deviation exceeds the limit, determining that the instruction execution is abnormal; For the instruction judged to be successfully executed, the mark in execution is cleared, the execution state feedback data containing the actual execution result is packaged, and the feedback data is transmitted back to the data preprocessing unit of the data acquisition module and the data analysis module in real time and is used for updating the real-time state library of the equipment; and if the command is still abnormal after retrying, generating an alarm event comprising the equipment address, the abnormal type and the time stamp, and pushing the alarm event to an AI fault diagnosis unit and an operation and maintenance alarm system of the data analysis module.

Description

AI tracker intelligent platform Technical Field The invention relates to the technical field of intelligent operation and maintenance of photovoltaics, in particular to an intelligent platform of an AI tracker. Background With the advancement of photovoltaic power plant marketization transactions, power plant operational objectives have shifted from maximizing power generation to maximizing power generation revenue. The tracking bracket can flexibly adjust the power generation curve according to the time-of-use electricity price, and becomes a key means for improving the asset value of the power station. Some photovoltaic monitoring systems exist at present, but the following outstanding technical problems still exist: In terms of control performance, the existing system mostly adopts a centralized architecture and a general communication protocol, so that second-level high-precision control of a mass tracker unit (NCU/TCU) is difficult to realize, and particularly in a region with complex network conditions or remote places, the problems of control delay and insufficient precision are remarkable, and the flexible adjustment capability of a power generation curve is limited. In terms of system deployment and data security, the existing scheme cannot meet different customer requirements, namely a large-scale ground power station has extremely high requirements on data main authority and real-time performance and needs localized deployment, and a distributed power station tends to be a cloud service with low cost and easy expansion. Existing systems often support a single mode, which makes it difficult to balance security and economy. In the operation and maintenance mode, most systems still rely on periodic inspection and post-maintenance, and lack of predictive maintenance capability based on artificial intelligence results in high operation and maintenance cost and slow fault response, thus causing unnecessary power generation loss. In terms of safety protection mechanisms, the existing system generally only has simple threshold alarming and passive protection functions, lacks an active risk avoiding strategy which is coordinated with the hardware depth of a tracker and is based on weather prediction, and has higher safety risk of equipment in severe weather. Therefore, an intelligent tracker management platform which can integrate AI diagnosis, support flexible deployment and has high-precision control and active safety protection capability is needed to comprehensively improve the benefits, safety and operation efficiency of a photovoltaic power station. Disclosure of Invention The invention aims to provide an intelligent platform of an AI tracker, which is used for solving the comprehensive problems of the photovoltaic tracker system in the aspects of control precision, operation and maintenance modes, safety protection and deployment flexibility in the prior art. In order to solve the technical problems, the invention specifically provides the following technical scheme: an AI tracker intelligent platform, comprising: The data acquisition module is used for acquiring operation data and environment data from an NCU or TCU unit of the photovoltaic tracker; the data analysis module is used for receiving the operation data and the environment data, performing fault diagnosis and state evaluation through an AI model and generating a health evaluation report; the control strategy generation module is used for generating a tracker control instruction according to the health evaluation report and the time-of-use electricity price data; and the remote control execution module is used for sending the tracker control instruction to the corresponding NCU or TCU unit to realize the adjustment of the tracker angle. As a preferred embodiment of the present invention, the data acquisition module specifically includes: The equipment data acquisition unit is used for acquiring running state data and electrical parameters including tracker angles, motor current and voltage, driving states and fault codes in real time through sensors built in the NCU or the TCU unit, and carrying out primary structural packaging on the acquired data; the environment data acquisition unit is used for connecting a meteorological station deployed by a photovoltaic power station or accessing an authoritative meteorological data service interface, and acquiring environment data and geographic information data comprising irradiance, environment temperature, wind speed and direction, rainfall and snowvolume and component temperature; The protocol adaptation unit is used for dynamically selecting or parallelly enabling communication protocols according to a power station deployment mode and network conditions, wherein the Zigbee protocol or the Lora protocol is preferentially enabled under a local area network environment to conduct low-delay and high-safety intranet communication, the 4G or 5G mobile network protocol is enabled when remote publi