CN-121989264-A - Grabbing state identification method and system based on photovoltaic panel paving robot

CN121989264ACN 121989264 ACN121989264 ACN 121989264ACN-121989264-A

Abstract

The invention relates to the technical field of control of photovoltaic panel paving equipment, and particularly discloses a grabbing state identification method and system based on a photovoltaic panel paving robot, wherein the method comprises the following steps: the method comprises the steps of collecting a pressure time sequence of a vacuum adsorption system of the robot in real time, calculating a negative duty ratio of the pressure difference sequence to obtain an attenuation judgment value, carrying out attenuation signal triggering judgment, extracting a historical adsorption signal group set if an attenuation signal is triggered, analyzing the relevance of adsorption pressure attenuation and attitude deviation, synchronously collecting core point coordinates of a photovoltaic panel, calculating attitude compensation quantity, evaluating compensation feasibility, calculating adsorption failure critical duration under the condition of linearity and nonlinearity if the compensation is feasible, predicting the current adsorption prediction residual duration by combining with historical installation duration, comparing the failure critical duration with the current adsorption prediction residual duration, carrying out normalized operation or emergency treatment as required, and improving the adsorption attenuation recognition accuracy and timeliness of the photovoltaic panel paving robot in grasping state recognition.

Inventors

ZHU XIAOQIANG
XU YUNWEI
XING RUI
FU JINJUN
YAN DANJIA
FU LIHUA
ZHOU GE

Assignees

中国能源建设集团浙江火电建设有限公司

Dates

Publication Date: 20260508
Application Date: 20260410

Claims (10)

1. A grabbing state identification method based on a photovoltaic panel paving robot is characterized by comprising the following steps of: collecting an adsorption pressure time sequence of a vacuum adsorption system of the photovoltaic panel paving robot in real time, and judging whether an adsorption attenuation signal is triggered or not through pressure characteristic trend time sequence analysis; If triggered, extracting a historical adsorption signal group set, carrying out adsorption attenuation influence analysis, judging whether adsorption attenuation causes adsorption posture offset, if so, collecting an adsorption installation coordinate group in real time, calculating posture compensation quantity of each core point of the photovoltaic panel, and carrying out feasibility analysis of grabbing identification compensation; if the method is feasible, based on the gesture compensation quantity of each core point of the photovoltaic panel, carrying out grabbing recognition compensation of the photovoltaic panel paving robot, and carrying out adsorption failure prediction analysis according to vacuum adsorption mechanics aiming at the photovoltaic panel paving robot subjected to grabbing recognition compensation to determine the critical duration of adsorption failure; And extracting time sequence data of the historical installation time, calculating the current residual adsorption prediction time, combining the adsorption failure critical time, judging the timeliness of the identification state, and carrying out corresponding identification response operation according to the judgment result.
2. The grabbing state identification method based on the photovoltaic panel paving robot, which is disclosed in claim 1, is characterized in that the process of judging whether to trigger the adsorption attenuation signal is as follows: the adsorption pressure time sequence refers to an adsorption pressure data set which is arranged in time sequence within the statistical time length from the start of grabbing the photovoltaic panel by the photovoltaic panel paving robot to the current time point; Based on any set of adjacent adsorption pressures in the adsorption pressure timing sequence: calculating the difference between the next adsorption pressure and the previous adsorption pressure to obtain an adsorption pressure difference; integrating all adsorption pressure difference values according to time sequence to form an adsorption pressure difference value sequence; counting the duty ratio of a negative value in the adsorption pressure difference sequence to obtain an adsorption attenuation judgment value; And triggering an adsorption attenuation signal if the adsorption attenuation judgment value is greater than or equal to the adsorption attenuation judgment threshold value.
3. The grabbing state identification method based on the photovoltaic board paving robot, which is disclosed by claim 1, is characterized in that the process of carrying out adsorption attenuation influence analysis and judging whether adsorption attenuation causes adsorption posture deviation is as follows: The triggering conditions of adsorption attenuation signals and adsorption attitude deviation signals of each pavement batch are extracted, and the adsorption attenuation signals and the adsorption attitude deviation signals are combined according to the same pavement batch to obtain a historical adsorption signal group; integrating the historical adsorption signal groups to obtain a historical adsorption signal group set; Marking paving batches triggering adsorption attenuation signals and adsorption posture offset signals simultaneously as adsorption attenuation influence batches based on a historical adsorption signal group set; Calculating the duty ratio of the adsorption attenuation influence batch in the total pavement batch to obtain an adsorption attenuation influence judgment value; And if the adsorption attenuation influence judgment value is larger than or equal to the adsorption attenuation influence judgment threshold value, describing that adsorption attenuation causes adsorption posture deviation.
4. The grabbing state recognition method based on the photovoltaic panel paving robot, which is disclosed by claim 1, is characterized in that the process of calculating the posture compensation quantity of each core point of the photovoltaic panel and carrying out the feasibility analysis of grabbing recognition compensation is as follows: the adsorption installation coordinate set refers to the combination of the coordinates of each core point of the current photovoltaic panel and the corresponding installation coordinates; the core point coordinates refer to four end point coordinates and geometric center origin coordinates; Calibrating unique identifiers of all core points of the photovoltaic panel, and respectively calculating vector difference values of coordinates of all core points of the current photovoltaic panel and corresponding installation coordinates according to the unique identifiers to obtain posture compensation amounts of all core points of the photovoltaic panel; If the gesture compensation quantity of each core point is in the mechanical arm kinematics constraint range, the grasping recognition compensation is judged to be feasible, and the coordinates of each core point of the current photovoltaic panel are adjusted according to the corresponding gesture compensation quantity.
5. The grabbing state identification method based on the photovoltaic panel paving robot, which is disclosed in claim 1, is characterized in that the process of determining the critical duration of adsorption failure is as follows: the dual adsorption failure prediction analysis refers to failure prediction analysis combining adsorption pressure attenuation trend and stress state after mechanical arm compensation; Setting a critical adsorption force standard, and calculating a variation coefficient of an adsorption pressure difference value based on an adsorption pressure difference value sequence to obtain an adsorption pressure fluctuation value; If the adsorption pressure fluctuation value is smaller than or equal to the adsorption pressure fluctuation threshold value, calculating the average change rate of the adsorption pressure value; Calculating the difference between the critical adsorption force standard and the current adsorption pressure, and carrying out ratio processing on the difference and the average change rate to obtain the adsorption failure critical time; If the adsorption pressure fluctuation value is larger than the adsorption pressure fluctuation threshold value, carrying out trend fitting by adopting a quadratic polynomial regression model and an exponential decay model nonlinear model based on the adsorption pressure time sequence, and selecting the nonlinear model corresponding to the maximum fitting goodness as an optimal model; and inverting the optimal model, and integrating to calculate the time accumulation from the current adsorption pressure to the critical adsorption force standard to obtain the adsorption failure critical duration.
6. The method for identifying the grabbing state based on the photovoltaic panel paving robot according to claim 5, wherein the process of calculating the average change rate of the adsorption pressure value is as follows: calculating the ratio of each adsorption pressure difference value to the data acquisition interval in the adsorption pressure difference value sequence to obtain the change rate of each adsorption pressure; and (5) taking an average value of all the change rates to obtain the average change rate of the adsorption pressure value.
7. The method for identifying the grabbing state based on the photovoltaic board laying robot according to claim 1 is characterized in that the process for calculating the current residual adsorption prediction time length is as follows: The time sequence data of the historical installation time length refers to the total flow operation time length from the completion of grabbing and posture adjustment of the photovoltaic panel to the accurate laying and positioning of the photovoltaic panel and the release of the sucker in each paving batch in the statistical time length; extracting the time length of the adsorption attenuation influence batch for installation, carrying out arithmetic mean processing, and marking as the current adsorption installation prediction time length; performing difference processing on the current adsorption installation predicted time length and the current performed time length to obtain a current adsorption predicted remaining time length; And the current performed time length is the product of the data acquisition interval and the data number in the adsorption pressure time sequence.
8. The grabbing state identification method based on the photovoltaic panel paving robot, which is disclosed by claim 1, is characterized in that the process of judging the identifying state time effectiveness is as follows: If the current adsorption installation prediction time length is greater than or equal to the adsorption failure critical time length, the fact that the safe adsorption time length cannot cover the time consumption of the operation is indicated, and the timeliness of the identification state is judged to be weak; If the current adsorption installation prediction time length is smaller than the adsorption failure critical time length, the safety window period is indicated to be time-consuming in the coverage operation, and the identification state is judged to be strong in timeliness.
9. The method for identifying the grabbing state based on the photovoltaic panel paving robot according to claim 1 is characterized in that the process of carrying out corresponding identification response operation according to the judging result is as follows: If the timeliness of the identification state is strong, no intervention is needed, a normalized installation operation flow is adopted, meanwhile, the change of the absorption force attenuation rate, the coordinate deviation of the core point of the photovoltaic panel and the absorption failure critical time length is dynamically identified in real time, if the timeliness is changed from strong to weak, the flow interruption is immediately triggered, and the emergency response mode is switched; And if the identification state is weak in timeliness, starting an emergency response mode, namely, adopting a nozzle to jet and clean foreign matters, then re-executing adsorption pressure acquisition and attenuation signal judgment, and if the adsorption attenuation signal is triggered again and the timeliness is still weak, contacting a maintenance department to carry out maintenance on internal devices of the robot.
10. A grabbing state recognition system based on a photovoltaic panel paving robot is characterized by comprising the following components: The attenuation signal judging module is used for collecting an adsorption pressure time sequence of the vacuum adsorption system of the photovoltaic panel paving robot in real time and judging whether to trigger an adsorption attenuation signal or not through pressure characteristic trend time sequence analysis; The gesture compensation evaluation module is used for extracting a historical adsorption signal set if triggering, carrying out adsorption attenuation influence analysis, judging whether adsorption attenuation causes adsorption gesture offset, acquiring an adsorption installation coordinate set in real time if the adsorption attenuation causes adsorption gesture offset, calculating gesture compensation quantity of each core point of the photovoltaic panel, and carrying out feasibility analysis of grabbing identification compensation; The failure duration prediction module is used for executing grabbing recognition compensation of the photovoltaic panel paving robot based on the gesture compensation quantity of each core point of the photovoltaic panel if the photovoltaic panel paving robot is feasible, and performing adsorption failure prediction analysis according to vacuum adsorption mechanics aiming at the photovoltaic panel paving robot subjected to grabbing recognition compensation to determine the adsorption failure critical duration; And the aging judging and responding module is used for extracting time sequence data of the historical installation time, calculating the current residual adsorption prediction time, judging the aging of the identification state by combining the adsorption failure critical time, and carrying out corresponding identification responding operation according to the judging result.

Description

Grabbing state identification method and system based on photovoltaic panel paving robot Technical Field The invention relates to the technical field of control of photovoltaic panel paving equipment, in particular to a grabbing state identification method and system based on a photovoltaic panel paving robot. Background Along with the continuous expansion of the construction scale of the photovoltaic power station, the photovoltaic development potential of the unused lands such as deserts and the like is widely excavated, the photovoltaic board paving robot is used as core equipment for replacing manual paving, and the accurate identification of the operation grabbing state is a key for guaranteeing the stable operation of the robot. However, in a photovoltaic panel paving construction site in a desert and other severe environments, the prior art has obvious defects that firstly, the existing grabbing state identification technology is not fully adapted to the desert and dust environment characteristics, the accuracy of identifying the adsorption pressure attenuation caused by the blocking of sand and dust by a sucker and the abrasion of a sealing element is low, misjudgment or missed judgment is easy to occur, and an initial abnormal state cannot be captured in time, secondly, a linkage analysis mechanism for the attenuation of the grabbing force of the sucker and the deviation of the gesture in the grabbing process is lacking, an accurate pre-judging model of the adsorption failure critical duration is not established, the safety risks of deviation, falling off, damage and the like of the photovoltaic panel are difficult to be regulated in advance, and the improvement of the installation efficiency of the photovoltaic panel cannot be realized, so that the operation reliability and the operation efficiency of a robot under the severe environment are restricted. Therefore, the invention provides a grabbing state identification method and system based on a photovoltaic panel paving robot. Disclosure of Invention The invention aims to provide a grabbing state identification method and system based on a photovoltaic panel paving robot, so as to solve the background problem. The aim of the invention can be achieved by the following technical scheme: A grabbing state identification method based on a photovoltaic panel paving robot comprises the following steps: collecting an adsorption pressure time sequence of a vacuum adsorption system of the photovoltaic panel paving robot in real time, and judging whether an adsorption attenuation signal is triggered or not through pressure characteristic trend time sequence analysis; If triggered, extracting a historical adsorption signal group set, carrying out adsorption attenuation influence analysis, judging whether adsorption attenuation causes adsorption posture offset, if so, collecting an adsorption installation coordinate group in real time, calculating posture compensation quantity of each core point of the photovoltaic panel, and carrying out feasibility analysis of grabbing identification compensation; if the method is feasible, based on the gesture compensation quantity of each core point of the photovoltaic panel, carrying out grabbing recognition compensation of the photovoltaic panel paving robot, and carrying out adsorption failure prediction analysis according to vacuum adsorption mechanics aiming at the photovoltaic panel paving robot subjected to grabbing recognition compensation to determine the critical duration of adsorption failure; And extracting time sequence data of the historical installation time, calculating the current residual adsorption prediction time, combining the adsorption failure critical time, judging the timeliness of the identification state, and carrying out corresponding identification response operation according to the judgment result. As a further scheme of the invention: Further, the process of judging whether to trigger the adsorption decay signal is as follows: the adsorption pressure time sequence refers to an adsorption pressure data set which is arranged in time sequence within the statistical time length from the start of grabbing the photovoltaic panel by the photovoltaic panel paving robot to the current time point; Based on any set of adjacent adsorption pressures in the adsorption pressure timing sequence: calculating the difference between the next adsorption pressure and the previous adsorption pressure to obtain an adsorption pressure difference; integrating all adsorption pressure difference values according to time sequence to form an adsorption pressure difference value sequence; counting the duty ratio of a negative value in the adsorption pressure difference sequence to obtain an adsorption attenuation judgment value; And triggering an adsorption attenuation signal if the adsorption attenuation judgment value is greater than or equal to the adsorption attenuation judgment threshold value. Further, the adsorp