Ground robot remote collaborative control method based on big data

Abstract

The application relates to a ground robot remote cooperative control method based on big data, and relates to the technical field of robot control, and comprises the following steps: acquiring a target area historical environment parameter distribution sequence based on big data, generating a robot control instruction at a specified position and sending the robot control instruction; indexing a historical environment parameter sequence of the specified position and configuring a moving speed; controlling the ground robot to move according to the instruction and the speed, analyzing a positioning error according to the historical environment parameter sequence, obtaining a specified position range after compensation; configuring an environment prediction resource according to the moving speed, predicting a prediction environment parameter set in the range, calculating an environment monitoring error degree, marking a parameter collected by the robot, and completing control. The application solves the problem that the existing ground robot remote control mode only depends on a preset route and a single signal, ignores the interference of air quality on remote data transmission, leads to large positioning deviation and high risk of parameter mis-collection, and improves the accuracy and reliability of environment monitoring.

Description

Technical Field

[0001] This application relates to the field of robot control, and in particular to a method for remote collaborative control of ground robots based on big data. Background Technology

[0002] With the intelligent development of ground environmental monitoring technology, the precise control of remote environmental parameter acquisition by robots has become an important foundation for regional environmental supervision and efficiency improvement. Currently, traditional ground robot environmental parameter acquisition mainly relies on fixed paths and single sensor identification, and does not consider the impact of air quality on remote data transmission, resulting in problems such as poor environmental adaptability, low positioning accuracy, and large data transmission errors.

[0003] Existing remote control methods rely solely on preset routes and single signal transmission for mobile control, ignoring the interference of air quality on remote data transmission. This results in significant deviations between the robot's actual positioning and the target monitoring point, increasing not only the risk of erroneous environmental parameter collection and the cost of manual calibration, but also failing to meet the accuracy requirements for parameter collection in intelligent ground environment monitoring. Summary of the Invention

[0004] To address the aforementioned technical challenges, this application provides a remote collaborative control method for ground robots based on big data. This method improves environmental adaptability and positioning accuracy, reduces data loss rate and transmission errors, and can meet the real-time and accurate acquisition requirements of multiple location parameters in complex environments.

[0005] This application discloses the following technical solution:

[0006] This application provides a method for remote collaborative control of ground robots based on big data, the method comprising:

[0007] Based on big data, the historical environmental parameter distribution sequence of the target area to be monitored is obtained, a specified location within the target area is selected, robot control commands are generated, and sent to the ground robot.

[0008] Index the historical environmental parameter sequence of the specified location to configure the movement speed;

[0009] The ground robot is controlled to move according to the robot control command and the moving speed. The positioning error is analyzed according to the historical environmental parameter sequence to obtain the positioning error parameter. The specified position is compensated to obtain the specified position range.

[0010] Based on the moving speed, configure environmental prediction resources, predict and obtain the set of predicted environmental parameters within the specified location range, calculate the environmental monitoring error degree, identify the location environmental parameters collected after the ground robot reaches the specified location, and complete the control.

[0011] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0012] This application proposes a remote collaborative control method for ground robots based on big data. Through historical environmental parameter analysis and dynamic error compensation, it achieves precise control of the robot's monitoring path. First, based on big data, the historical environmental parameter distribution sequence of the target area is obtained, the average historical environmental parameters at each location coordinate are calculated, and the most polluted designated locations are selected, generating robot control commands. Then, the historical parameter sequence of the designated location is indexed, and the movement speed is configured by comparing it with standard parameters. When controlling the robot to move according to commands, the latest historical environmental parameters are selected and input into a positioning error classification table to obtain positioning error parameters, which are then compensated for at the designated locations to determine the actual monitoring range. Simultaneously, environmental prediction resources are configured based on the movement speed, and a set of environmental parameters is predicted through machine learning to calculate the monitoring error degree to identify the collected parameters. When the measured value exceeds the error range, control commands such as resampling or sensor calibration are automatically triggered.

[0013] The technical solution of this application solves the problems of traditional control methods ignoring air quality interference and large positioning deviations by integrating multi-dimensional information such as historical environmental data, real-time positioning errors and prediction parameters, thereby achieving precise and intelligent environmental monitoring of ground robots and effectively improving the reliability of data collection and the efficiency of environmental supervision. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 A flowchart illustrating the remote collaborative control method for ground robots based on big data provided in this application embodiment;

[0016] Figure 2 This is a flowchart illustrating the ground robot positioning error analysis and position compensation provided in an embodiment of this application. Detailed Implementation

[0017] This application provides a remote collaborative control method for ground robots based on big data, which solves the technical problem that the control method in the prior art relies only on fixed paths and single recognition, without considering air quality interference, resulting in large errors in the positioning and parameter transmission of ground robots.

[0018] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0019] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0020] In the description of this application, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0021] Examples, as shown in the appendix Figure 1 As shown, this application provides a remote collaborative control method for ground robots based on big data, the method comprising the following steps:

[0022] S110: Based on big data, obtain the historical environmental parameter distribution sequence of the target area to be monitored, select a specified location within the target area, generate robot control commands, and send them to the ground robot;

[0023] In this embodiment of the application, in the environmental monitoring scenario of factory areas and other regions, in order to accurately determine the location of key environmental parameters to be collected, it is necessary to use big data to mine historical data in order to achieve efficient and accurate regional environmental monitoring planning.

[0024] Specifically, the first step is to collect historical environmental parameters of the target area, including air pollutant concentrations, and construct a distribution sequence of historical environmental parameters to facilitate the analysis of data patterns from both temporal and spatial dimensions.

[0025] Furthermore, the average historical environmental parameters corresponding to the coordinates of each location within the area are calculated. By comparing the data, the location with the highest average pollution level is selected as the designated monitoring point.

[0026] Finally, based on the designated location, precise robot movement and data collection instructions are generated and sent to guide the ground robot to conduct environmental parameter collection. By using big data to drive the selection of designated locations, the targeting and efficiency of regional environmental monitoring are improved.

[0027] Step S110 in the method provided in this application embodiment includes:

[0028] Based on big data, the historical environmental parameter distribution sequence of the target area to be monitored is obtained, where each environmental parameter includes air pollution parameters;

[0029] Calculate the average historical environmental parameters of all location coordinates within the target area;

[0030] Select the location coordinates with the highest average historical environmental parameters as the specified location;

[0031] Based on the specified location, robot control commands are generated and sent to the ground robot.

[0032] In this embodiment of the application, in environmental monitoring scenarios such as factory areas, in order to accurately determine the most polluted locations where key environmental parameters need to be collected, monitoring paths are planned by mining historical data using big data.

[0033] Specifically, by deploying 50 air quality sensors in multiple high-pollution areas such as the factory's welding workshop and chemical warehouse, air pollution parameters for the past 365 days are continuously collected at a frequency of 10 minutes per time, forming a historical environmental parameter distribution sequence that includes timestamps accurate to the minute and spatial coordinates with an error of ≤0.5 meters.

[0034] Each environmental parameter corresponds to a specific air pollution parameter. Through long-term data collection from 50 monitoring points across the entire region, a historical environmental parameter distribution sequence covering the target area was constructed, providing a foundation for subsequent data analysis.

[0035] Furthermore, the target area is divided into 2000 grid cells with a grid precision of 5m×5m, and each grid cell corresponds to a unique location coordinate. For each grid cell, all historical environmental parameter data within its coverage area are extracted, and the average historical environmental parameter is calculated by summing them and dividing by the total amount of data (i.e., average historical environmental parameter = Σ historical environmental parameter value / number of data).

[0036] For example, a grid cell collected data at a frequency of 10 minutes per data point over the past 365 days, resulting in a total data volume of 365 × 24 × 6 = 52,560 data points. The air pollution parameters of these 52,560 data points were summed one by one and divided by 52,560 to obtain the average historical environmental parameters of the grid cell.

[0037] Furthermore, by iterating and comparing the average historical environmental parameters of 2000 grid cells, the location coordinates with the largest values ​​were selected as the designated locations for subsequent monitoring.

[0038] Furthermore, based on the acquired specified location, robot control commands are generated and sent to the ground robot to guide the robot in performing precise movement and monitoring tasks.

[0039] The step of "generating robot control instructions based on the specified position" in the method provided in this application embodiment includes:

[0040] Obtain the real-time location of the ground robot and an electronic map of the target area;

[0041] Within the electronic map of the area, the real-time location and the movement route to the specified location are generated, and robot control commands are generated to control the robot's movement.

[0042] In this embodiment of the application, when generating robot control commands, it is necessary to first obtain the real-time location of the ground robot and the electronic map of the target area to provide data support for the planning of the movement route.

[0043] Specifically, the GPS positioning system on the ground robot is used to obtain the robot's current position coordinates in real time. At the same time, an electronic map of the area containing obstacle distribution and road signs is loaded from the server. The map accuracy must meet the robot's movement control requirements (such as 1 meter × 1 meter grid accuracy).

[0044] Furthermore, based on the robot's real-time position and the designated position, when planning the movement route within the electronic map, geometric collision detection is first performed by connecting the coordinate range of obstacles in the electronic map (such as the coordinate range from the lower left corner to the upper right corner of a rectangular obstacle) with the straight line from the real-time position to the designated position to determine whether there are obstacles between the two positions.

[0045] If there are no obstacles between two locations, a straight-line movement route is generated directly; if there are obstacles, the movement is based on the grid coordinates of the electronic map and uses the "axis-bypass" method, prioritizing movement along the X-axis or Y-axis to the edge of the obstacle before turning to the designated location.

[0046] For example, if the real-time position of the ground robot is (100, 100) and the specified position is (300, 200), and an obstacle occupies the area from (200, 150) to (250, 180), the robot's detour route is from (100, 100) to (200, 100) to (200, 200) to (300, 200). This process determines the robot's movement path through coordinate comparison.

[0047] Furthermore, robot control commands are generated based on the planned movement route and sent to the ground robot. These commands include information such as the movement path and speed.

[0048] This process uses coordinate analysis from an electronic map and obstacle collision detection to generate robot control commands, ensuring that the robot moves precisely to the designated location along the planned path.

[0049] S120: Index the location history environmental parameter sequence of the specified location and configure the movement speed;

[0050] In this embodiment of the application, in environmental monitoring scenarios such as factory areas, in order to enable the ground robot to adapt to the environmental conditions of different polluted areas, it is necessary to dynamically configure the movement speed according to the historical environmental parameters of the specified location in order to optimize monitoring efficiency and data acquisition accuracy.

[0051] Specifically, the system first indexes the historical environmental parameter sequence for a specified location, which contains environmental parameter data such as air pollutant concentrations at that location over past time periods. Then, it calculates the average historical environmental parameter for that location, thus reflecting the overall pollution level at that location.

[0052] Furthermore, the ratio of the average historical environmental parameters to the standard environmental parameters (i.e., average historical environmental parameters / standard environmental parameters) is calculated, and the preset movement speed is configured based on this ratio to obtain the robot's movement speed.

[0053] This step allows the robot's movement speed to be dynamically configured based on historical environmental parameters at a specified location, enabling the robot to better adapt to different environmental conditions and improving the flexibility and adaptability of ground robots in environmental monitoring tasks.

[0054] Step S120 in the method provided in this application embodiment includes:

[0055] Index the location history environmental parameter sequence of the specified location and calculate the mean to obtain the average location history environmental parameter;

[0056] The ratio of the average historical environmental parameters to the standard environmental parameters is calculated, and the preset movement speed is configured to obtain the movement speed.

[0057] In this embodiment of the application, in order to enable the ground robot to adaptively adjust its movement speed according to the environmental conditions of the specified location, dynamic speed configuration needs to be achieved based on the quantitative analysis of historical environmental parameters in order to balance the robot's monitoring accuracy and movement efficiency.

[0058] Specifically, the system first indexes the location's historical environmental parameter sequence, which contains air pollution parameters collected at a frequency of 10 minutes per day over the past 365 days, with the data accompanied by minute-level timestamps and spatial coordinates ≤0.5 meters.

[0059] Furthermore, the mean of all environmental parameters in the sequence is calculated to obtain the average location historical environmental parameters.

[0060] For example, at a specified location, its historical parameter sequence contains 365×24×6=52560 data points. The environmental parameter values ​​of these data are summed one by one and then divided by 52560 to obtain the average historical environmental parameter of the location (e.g., 80).

[0061] Furthermore, the ratio of the average historical environmental parameter to the standard environmental parameter (i.e., average historical environmental parameter / standard environmental parameter) is calculated based on the reference value of the standard environmental parameter for the clean area, and the preset movement speed is configured based on this ratio.

[0062] When the ratio of the two is greater than 1, it indicates that the environmental parameter value is higher than the standard. The speed will be reduced by multiples of the ratio to slow down the robot in high-pollution areas and extend the data collection time. If the ratio is less than 1, the speed will be increased by multiples of the ratio to accelerate in low-pollution areas and improve monitoring efficiency.

[0063] For example, if the average historical environmental parameter of a specified location is 80 and the standard environmental parameter is 35, then the ratio of the two is 80 / 35≈2.29. Since 2.29>1, the ground robot needs to reduce its moving speed to 0.5 / 2.29≈0.218m / s according to this ratio to ensure that the robot slows down its movement in highly polluted areas, increases the data collection time, and thus improves the accuracy of mobile monitoring.

[0064] Conversely, if the average historical environmental parameter of a specified location is 28, then the ratio of the two is 28 / 35=0.8. Since 0.8<1, the preset moving speed of the ground robot is increased from 0.5m / s to 0.5 / 0.8=0.625m / s, so that it can move faster in low-pollution areas and improve the overall monitoring efficiency.

[0065] By configuring the speed of the ground robot based on the above environmental parameter values, the robot's movement speed is negatively correlated with the environmental parameter values. This ensures fine sampling in areas with high parameter values, improves overall monitoring efficiency, and solves the problem of poor environmental adaptability in the traditional fixed speed mode.

[0066] S130: Control the ground robot to move according to the robot control command and the moving speed, perform positioning error analysis based on the historical environmental parameter sequence, obtain positioning error parameters, compensate for the specified position, and obtain the specified position range;

[0067] In this embodiment, when the ground robot moves to a designated location, air pollutants can interfere with the accuracy of the positioning signal, causing positioning deviation. To ensure that the robot can accurately reach the target area to collect data, it is necessary to perform positioning error analysis based on historical environmental parameter sequences and determine a reasonable range for the designated location through dynamic compensation.

[0068] Specifically, the ground robot is controlled to move in the target area according to control commands and speed, while the latest historical environmental parameters are selected from the historical environmental parameter sequence.

[0069] Furthermore, the latest historical environmental parameters are input into the positioning error classification table. This table maps the sample environmental parameter set and the sample positioning error parameter set according to the latest historical environmental parameters, and outputs the positioning error parameters through classification retrieval.

[0070] Finally, the positioning error parameter is used to compensate for the location. In this way, even if there are errors in the robot's positioning, valid data can still be collected within this range, mitigating the impact of air pollution on positioning accuracy and improving the reliability of data acquisition.

[0071] This step involves analyzing historical environmental parameters to obtain positioning error parameters, and then using these parameters to dynamically compensate for the location at a specified point, forming a reasonable range to ensure the robot's positioning accuracy and data collection effectiveness in polluted environments.

[0072] As attached Figure 2 As shown, step S130 in the method provided in this application embodiment includes:

[0073] The ground robot is controlled to move according to the robot control command and the moving speed, and the latest historical environmental parameter in the historical environmental parameter sequence is selected;

[0074] The latest historical environmental parameters are input into the positioning error classification table, and the positioning error parameters are obtained by classification output. The positioning error classification table includes a set of mapped sample environmental parameters and a set of sample positioning error parameters. The positioning error parameters include positioning error distance.

[0075] The specified location is compensated using the positioning error parameters to obtain the specified location range.

[0076] In this embodiment, when a ground robot performs a mobile task, airborne pollution particles can interfere with the transmission of positioning signals, causing a deviation between the actual position and the target position. To eliminate the impact of environmental pollution on positioning accuracy, error analysis needs to be conducted based on real-time data of historical environmental parameters, and a reliable monitoring range needs to be obtained through dynamic compensation.

[0077] Specifically, the ground robot moves to the designated location according to the generated control commands (including movement path and movement speed), and at the same time extracts the latest environmental parameter value from the historical environmental parameter sequence. For example, the environmental parameter value collected at the most recent moment is 150, which represents the current pollution level.

[0078] Furthermore, the latest historical environmental parameter values ​​are input into the positioning error classification table to output the corresponding positioning error parameters.

[0079] The positioning error classification table includes a set of mapped sample environmental parameters and a set of sample positioning error parameters (including positioning error distance). The positioning error parameters include the positioning error distance. For example, a sample environmental parameter value of 100 corresponds to a positioning error distance of 2 meters, 150 corresponds to 3 meters, and 200 corresponds to 5 meters.

[0080] By matching the latest historical environmental parameter values ​​with the sample environmental parameter set in the classification table, the corresponding positioning error parameters can be output, thus providing a basis for subsequent compensation processing of the specified location to obtain the specified location range. For example, if the latest environmental parameter value is 150, the positioning error parameter can be directly output as 3 meters after looking up the table (i.e., the positioning error distance is 3 meters).

[0081] Furthermore, to avoid the impact of positioning errors on the accuracy of environmental parameter acquisition, compensation processing needs to be performed on the specified location based on the positioning error distance.

[0082] Specifically, the positioning error distance in the positioning error parameters is used. With the specified location as the center and the positioning error distance as the radius, a circular range of the specified location is obtained. This circular area is the specified location range after compensation.

[0083] By using the dynamic compensation method described above, the positioning error caused by factors such as air pollution during the movement of the ground robot can be effectively eliminated. As long as it reaches the compensated range, it is determined to have reached the target position, thereby ensuring that the ground robot can accurately collect environmental parameters near the designated location and reduce data collection deviation caused by positioning error.

[0084] S140: Configure environmental prediction resources according to the moving speed, predict and obtain the predicted environmental parameter set within the specified location range, calculate the environmental monitoring error degree, identify the location environmental parameters collected after the ground robot reaches the specified location, and complete the control.

[0085] In this embodiment of the application, when the ground robot performs environmental monitoring tasks, in order to balance monitoring efficiency and data accuracy, it is necessary to dynamically configure environmental prediction resources according to the robot's real-time moving speed to achieve accurate prediction and error identification of environmental parameters within a specified location range.

[0086] Specifically, the maximum moving speed of the ground robot is first obtained. The environmental prediction resource coefficient is obtained by calculating "1 minus the ratio of the moving speed to the maximum moving speed". The magnitude of this coefficient is negatively correlated with the moving speed.

[0087] Furthermore, from the pre-trained sequence of environmental parameter predictors, a corresponding number of predictors are randomly selected according to the proportion of environmental prediction resource coefficients to match the resource input requirements under the current movement speed.

[0088] Meanwhile, within the historical environmental parameter distribution sequence, a set of historical environmental parameters for all location coordinates within a specified location range is selected as input data for the prediction model.

[0089] Furthermore, the selected historical parameter set sequence is input into the selected predictor, the output results of each predictor are aggregated and their mean is calculated to obtain the corresponding prediction environment parameters, and then integrated to generate a prediction environment parameter set.

[0090] Finally, based on the predicted environmental parameter set, the maximum error range of the predicted environmental parameters is determined by calculating the deviation range between each predicted environmental parameter and the historical environmental parameters, thus obtaining the environmental monitoring error degree. When the robot reaches the designated location range, the collected actual environmental parameters are combined with this error degree to identify the location environmental parameters, completing the robot's monitoring and control.

[0091] This process achieves synergistic optimization of efficiency and accuracy by dynamically matching movement speed with predicted resources, thus solving the problem of poor adaptability to the monitoring environment caused by fixed resource configuration in traditional control methods.

[0092] Step S140 in the method provided in this application embodiment includes:

[0093] Obtain the maximum moving speed of the ground robot, and calculate 1 minus the ratio of the moving speed to the maximum moving speed as the environmental prediction resource coefficient;

[0094] Obtain the environmental parameter predictor sequence;

[0095] Within the historical environmental parameter distribution sequence, multiple historical location environmental parameter set sequences are obtained by filtering all location coordinates within the specified location range;

[0096] An environmental parameter predictor that randomly selects the proportion of the environmental predicted resource coefficient inputs multiple historical location environmental parameter set sequences within a specified location range, obtains multiple prediction result sets and calculates the mean, and obtains multiple predicted environmental parameters as a predicted environmental parameter set;

[0097] Based on the predicted environmental parameter set and the calculated environmental monitoring error degree, the location environmental parameters collected after the ground robot reaches the designated location are identified, and control is completed.

[0098] In this embodiment of the application, in order to achieve accurate prediction of environmental parameters, prediction resources need to be dynamically configured according to the actual moving speed of the robot in order to accurately identify the parameters collected by the robot and ensure the reliability and effectiveness of the monitoring data.

[0099] Specifically, the maximum moving speed of the ground robot is first obtained through the robot control system, and then the environmental prediction resource coefficient of the robot is calculated by combining the calculation formula "environmental prediction resource coefficient = 1 - moving speed / maximum moving speed".

[0100] The environmental prediction resource coefficient is negatively correlated with the robot's movement speed. The larger the coefficient, the slower the robot moves, and the more environmental prediction resources (number of prediction models) can be allocated to improve prediction accuracy. The smaller the coefficient, the faster the robot moves, and the fewer environmental prediction resources (number of prediction models) need to be allocated to improve prediction efficiency.

[0101] For example, if the robot's maximum moving speed is 1 m / s and the current moving speed is 0.5 m / s, then the environmental prediction resource coefficient = 1 - 0.5 / 1 = 0.5, indicating that 50% of the prediction model numbers need to be called for parameter prediction; if the current moving speed increases to 0.8 m / s, then the environmental prediction resource coefficient = 1 - 0.8 / 1 = 0.2, only 20% of the prediction model numbers need to be called, thus improving computational efficiency while ensuring basic accuracy.

[0102] The step of "obtaining the environmental parameter predictor sequence" in the method provided in this application embodiment includes:

[0103] Based on environmental big data, a set of environmental parameter sequences from historical locations of samples is collected, and environmental parameters from different historical locations of samples are collected and labeled as a set of predicted environmental parameters for the samples.

[0104] Based on machine learning, construct multiple environmental parameter predictors;

[0105] Each time, a portion of sample data is randomly selected from the set of historical location environmental parameters and the set of predicted environmental parameters of the samples. Supervised training is performed on each environmental parameter predictor until convergence, resulting in a sequence of trained environmental parameter predictors.

[0106] In this embodiment of the application, in order to achieve accurate prediction of environmental parameters, it is necessary to construct and train a sequence of environmental parameter predictors so that they have the ability to accurately predict environmental parameters at a specified location.

[0107] Specifically, firstly, relying on the environmental big data collection system, historical environmental parameters of each location coordinate are continuously collected in the target area at a preset sampling frequency (e.g., 10 minutes / time), forming a set of sample historical location environmental parameter sequences.

[0108] The sample's historical location environmental parameter set sequence includes parameter sequences for each location in the time dimension (such as the environmental parameter value sequence for the past 365 days), and each parameter sequence is accompanied by a precise timestamp and spatial coordinates.

[0109] Simultaneously, for each sample's historical location environmental parameter set sequence, the environmental parameter values ​​following that sequence are collected and labeled as the sample's predicted environmental parameter set.

[0110] For example, the environmental parameter sequence of a certain location coordinate from January 1 to January 31, 2025 is the sample history sequence, and the environmental parameters on February 1 are collected as the corresponding predicted label value to form a "historical sequence-predicted value" sample pair.

[0111] Furthermore, multiple environmental parameter predictors are constructed based on machine learning.

[0112] Specifically, a random forest algorithm is employed, integrating the predictive capabilities of multiple decision trees to learn patterns in historical environmental parameters. First, the collected historical environmental parameter data is divided into training, validation, and test sets in a 7:2:1 ratio to ensure the data distribution reflects parameter characteristics under different pollution levels. For example, 3500 training sets, 1000 validation sets, and 500 test sets are extracted from 5000 data sets.

[0113] Furthermore, when training the random forest model using the training set, 80% of the sample data is randomly selected from the set of historical location environment parameters and the set of predicted environment parameters of the samples each time, and then input into the predictor.

[0114] Meanwhile, historical environmental parameter values ​​in the time series (such as parameter values ​​in the past 72 hours or every 10 minutes) are used as input features, and the environmental parameter values ​​at the next moment are used as output labels. The parameter change patterns are learned through the automatic splitting mechanism of the decision tree.

[0115] During training, the initial number of decision trees was set to 50, and the model performance was evaluated using a validation set every 10 trees added. After the first round of training, the average error between the predicted environmental parameter values ​​and the actual values ​​on the validation set was 0.12; when the number of decision trees increased to 100, the error decreased to 0.06, indicating that the model's ability to fit the changing trends of environmental parameters gradually improved.

[0116] Meanwhile, to avoid overfitting, an early stopping mechanism is implemented. Training is stopped when the error on the validation set no longer decreases for five consecutive rounds (error fluctuation is less than 0.003), indicating that the model has converged. For example, if the validation set error stabilizes at 0.05 during the 30th training round and has not decreased for five consecutive rounds, the model is considered converged, and training is stopped.

[0117] Ultimately, the trained environmental parameter predictor sequence can receive historical environmental parameter sequences from a specified location and output predicted values ​​of future environmental parameters, providing a quantitative basis for calculating environmental monitoring error.

[0118] Furthermore, after the environmental parameter predictor is constructed, the historical environmental parameter sequence corresponding to each coordinate point within the specified location range (such as the parameters of each point within a radius of 5 meters centered on the specified location) is input into the trained predictor to obtain the predicted environmental parameter value for each point.

[0119] For example, in a 10m × 10m monitoring area, among the 121 grid points generated at 1m intervals, there are 118 valid points. After processing by the predictor, each point outputs a predicted value of environmental parameters for the next hour, forming a prediction set containing 118 values.

[0120] Furthermore, an arithmetic mean is calculated for all the predicted values ​​in this set to obtain the predicted environmental parameter set for the region.

[0121] For example, if the sum of the predicted values ​​for 118 locations is 9440, then the average predicted value for the region is 9440 / 118=80, which means that the average predicted value of environmental parameters for the region in the next hour is 80.

[0122] Furthermore, the environmental monitoring error degree is calculated based on the predicted environmental parameter set to identify the actual environmental parameters collected by the ground robot.

[0123] The method provided in this application includes the following steps: "Based on the predicted environmental parameter set and calculating the environmental monitoring error degree, identifying the location environmental parameters collected after the ground robot reaches the designated location, and completing the control."

[0124] Based on the predicted environmental parameter set, the maximum error amplitude of the predicted environmental parameters is calculated to obtain the environmental monitoring error degree;

[0125] After the ground robot reaches the designated location, it acquires the collected location environmental parameters, identifies the location environmental parameters using the environmental monitoring error rate, and completes the control.

[0126] In this embodiment of the application, in order to accurately identify the environmental parameters collected by the ground robot, it is necessary to calculate the error range based on the predicted environmental parameter set in order to determine the monitoring error degree and complete the parameter identification.

[0127] Specifically, the first step is to analyze all predicted values ​​in the predicted environmental parameter set. By calculating the range of deviations between the predicted values ​​and historical true values, the maximum error margin is determined. That is, the true parameter values ​​corresponding to the prediction time period are selected from the historical environmental parameters, the absolute difference between each predicted value and the true value is calculated, and the largest difference is taken as the maximum error margin.

[0128] For example, if the predicted environmental parameter set contains 118 predicted values, the average value of the corresponding historical true values ​​is 75, and the largest difference between each predicted value and 75 is 7, then the maximum error range is 7, and the environmental monitoring error degree is ±7.

[0129] Furthermore, after the ground robot reaches the designated location, it acquires the collected location environmental parameters and uses environmental monitoring error rate to identify the reliability range of the data.

[0130] Specifically, the actual environmental parameter values ​​collected by the robot are first obtained, such as a parameter value of 80 collected at a certain moment, and the calculated environmental monitoring error degree ±7 is called. By combining the measured value and the error degree, the value is labeled in the format of "measured value ± error degree", that is, the labeling result "80±7" is generated.

[0131] This labeling result visually presents the range of data fluctuations, enabling environmental analysts to quickly understand the reliability of measured data when conducting environmental parameter analysis. For example, "80±7" indicates that the actual value of the environmental parameter may fluctuate between 73 and 87, providing a clear error reference for data interpretation.

[0132] Ultimately, the obtained identification results are used to control the ground robot, ensuring the reliability of monitoring data and the accuracy of environmental supervision.

[0133] For example, if the identification result is "80±7", and the actual parameter value collected by the ground robot is 82, which is within the error range of "73~87", it indicates that the data collection of the target area is effective and no additional operation is required.

[0134] Conversely, if the actual parameter value collected by the ground robot is 90, exceeding the upper limit of 87, the robot will automatically trigger a resampling command, and the robot will re-collect environmental parameters at that location until the data returns to the error range of 73~87.

[0135] In addition, if three consecutive sampling values ​​exceed the error range, the robot sensor is deemed abnormal. A calibration command will be sent to the robot to suspend its data acquisition task and arrange for professional personnel to check whether the sensor needs to be recalibrated, in order to further ensure the validity of the monitoring data and the accuracy of environmental supervision.

[0136] The embodiments of this application, through the specific implementation methods described above, achieve the following technical effects:

[0137] This application proposes a remote collaborative control method for ground robots based on big data. First, historical environmental parameters of the target area are acquired using big data, and the most polluted locations are selected to generate robot control commands, ensuring targeted monitoring. Second, historical parameters of the specified locations are indexed, and their ratio to standard parameters is calculated. The robot's movement speed is dynamically configured: if environmental parameters are higher than the standard, the robot decelerates for finer sampling; conversely, it accelerates to improve efficiency, addressing the poor adaptability of traditional fixed-speed methods. Next, the robot is controlled to move according to commands, and the latest historical parameters are input into a positioning error classification table to obtain and compensate for the error distance. This distance is used as the radius to generate a specified location range, eliminating interference from air pollution on positioning. Finally, an environmental prediction resource coefficient is calculated based on the movement speed, and prediction resources are configured. A predictor trained by machine learning predicts environmental parameters within the specified range, and the maximum error amplitude is calculated to obtain the monitoring error degree. Upon arrival, the robot combines the collected parameters with the error degree and identifies it as "measured value ± error degree." If the value exceeds the range, resampling or sensor calibration is triggered, achieving closed-loop control.

[0138] The method provided in this application addresses the problems of traditional ground robot control methods neglecting air quality interference and having large positioning deviations. Through the steps of "environmental parameter driving - dynamic configuration of moving speed - error distance compensation - intelligent identification", it improves the accuracy and intelligence level of ground robot environmental monitoring and provides reliable data support for environmental supervision.

[0139] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0140] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0141] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application intends to include such modifications and variations.

Claims

1. A remote collaborative control method for ground robots based on big data, characterized in that, The method includes: based on big data, obtaining the historical environmental parameter distribution sequence of the target area to be monitored, selecting a specified location within the target area, generating robot control commands, and sending them to the ground robot; Index the historical environmental parameter sequence of the specified location to configure the movement speed: The system indexes the sequence of historical environmental parameters for a specified location, calculates the mean to obtain the average historical environmental parameters, calculates the ratio of the average historical environmental parameters to the standard environmental parameters, and configures the preset movement speed based on this ratio. When the ratio is greater than 1, the speed is reduced by a multiple of the ratio; when the ratio is less than 1, the speed is increased by a multiple of the ratio. The ground robot is controlled to move according to the robot control command and the moving speed. The positioning error is analyzed according to the historical environmental parameter sequence to obtain the positioning error parameter. The specified position is compensated to obtain the specified position range. Configure environmental prediction resources according to the moving speed, predict and obtain the predicted environmental parameter set within the specified location range, calculate the environmental monitoring error degree, identify the location environmental parameters collected after the ground robot reaches the specified location, and complete the control. Based on the moving speed, environmental prediction resources are configured to predict and obtain a set of predicted environmental parameters within the specified location range, and the environmental monitoring error degree is calculated. The location environmental parameters collected after the ground robot reaches the specified location are then identified, and control is completed, including: Obtain the maximum moving speed of the ground robot, and calculate 1 minus the ratio of the moving speed to the maximum moving speed as the environmental prediction resource coefficient; Obtain the environmental parameter predictor sequence; Within the historical environmental parameter distribution sequence, multiple historical location environmental parameter set sequences are obtained by filtering all location coordinates within the specified location range; Using the environmental prediction resource coefficient as the calling ratio, randomly select a number of environmental parameter predictors corresponding to the calling ratio within the environmental parameter predictor sequence, input multiple historical location environmental parameter set sequences within a specified location range, obtain multiple prediction result sets and calculate the average value to obtain multiple predicted environmental parameters, which are used as the predicted environmental parameter set. Based on the predicted environmental parameter set and the calculated environmental monitoring error degree, the location environmental parameters collected after the ground robot reaches the designated location are identified to complete the control. Based on the predicted environmental parameter set and the calculated environmental monitoring error degree, the location environmental parameters collected after the ground robot reaches the designated location are identified, and control is completed, including: Based on the predicted environmental parameter set, the maximum error amplitude of the predicted environmental parameters is calculated to obtain the environmental monitoring error degree; After the ground robot reaches the designated location, it acquires the collected location environmental parameters, uses the environmental monitoring error rate to identify the location environmental parameters, and completes the control. After the ground robot reaches the designated location, it acquires the collected location environmental parameters, identifies the location environmental parameters using the environmental monitoring error degree, and completes the control. This includes: first, acquiring the actual environmental parameter values ​​collected by the robot, and simultaneously calling the calculated environmental monitoring error degree, and then identifying them using the format of measured value ± error degree to generate an identification result.

2. The remote collaborative control method for ground robots based on big data according to claim 1, characterized in that, Based on big data, the system obtains the historical environmental parameter distribution sequence of the target area to be monitored, selects a specific location within the target area, generates robot control commands, and sends them to the ground robot, including: Based on big data, the historical environmental parameter distribution sequence of the target area to be monitored is obtained, where each environmental parameter includes air pollution parameters; Calculate the average historical environmental parameters of all location coordinates within the target area; Select the location coordinates with the highest average historical environmental parameters as the specified location; Based on the specified location, robot control commands are generated and sent to the ground robot.

3. The remote collaborative control method for ground robots based on big data according to claim 2, characterized in that, Based on the specified location, generate robot control instructions, including: Obtain the real-time location of the ground robot and an electronic map of the target area; Within the electronic map of the area, the real-time location and the movement route to the specified location are generated, and robot control commands are generated to control the robot's movement.

4. The remote collaborative control method for ground robots based on big data according to claim 1, characterized in that, Controlling the ground robot to move according to the robot control commands and movement speed, performing positioning error analysis based on the historical environmental parameter sequence to obtain positioning error parameters, compensating for the specified location, and obtaining the specified location range, including: The ground robot is controlled to move according to the robot control command and the moving speed, and the latest historical environmental parameter in the historical environmental parameter sequence is selected; The latest historical environmental parameters are input into the positioning error classification table, and the positioning error parameters are obtained by classification output. The positioning error classification table includes a set of mapped sample environmental parameters and a set of sample positioning error parameters. The positioning error parameters include positioning error distance. The specified location is compensated using the positioning error parameters to obtain the specified location range.

5. The remote collaborative control method for ground robots based on big data according to claim 1, characterized in that, Obtain the environmental parameter predictor sequence, including: Based on environmental big data, a set of environmental parameter sequences from historical locations of samples is collected, and environmental parameters from different historical locations of samples are collected and labeled as a set of predicted environmental parameters for the samples. Based on machine learning, construct multiple environmental parameter predictors; Each time, a portion of sample data is randomly selected from the set of historical location environmental parameters and the set of predicted environmental parameters of the samples. Supervised training is performed on each environmental parameter predictor until convergence, resulting in a sequence of trained environmental parameter predictors.

Images