A method and device for data collection of automatic guided transport vehicles for mining

Abstract

The present invention belongs to the technical field of data collection. The present invention provides a method and device for data collection of an automatic guided vehicle for mining, including: identifying the target data collection type according to the collection integrity of the target data, classifying the target data by analyzing the target data collection type, and performing stability analysis on the occurrence of historical collection cycles of collection failures corresponding to the classified risk data, determining the predicted cycle of risk data collection failure to facilitate optimization of subsequent data collection, comparing the predicted cycle of risk data collection failure with the current collection cycle, and judging whether deduplication optimization of risk data collection is required before the automatic guided vehicle for mining sends risk data to a distributed message queue. If necessary, in the deduplication operation before the risk data is sent, the deduplication mechanism is optimized, which is beneficial to improving the integrity of data collection of the automatic guided vehicle for mining in the current collection cycle.

Description

Technical Field

[0001] The invention belongs to the technical field of data collection, and in particular to a method and a device for collecting data of an automatic guided transport vehicle for mining. Background Art

[0002] During the data collection process of the automatic guided vehicle for mining, the automatic guided vehicle for mining will send data in batches. These data are transmitted based on the user datagram protocol (such as UDP), and the receiving end will receive these data. Since the automatic guided vehicle for mining will send the same data multiple times, it is necessary to deduplicate the received data. The deduplicated data will be sent to the distributed message queue. The distributed message queue has a very high throughput and can collect data to the data center in real time. However, in the above-mentioned prior art, the main problem is that during the data collection process of the automatic guided vehicle for mining, the same data will be deduplicated, resulting in the number of copies of data being sent and collected being 1. It is very likely that the data collection will be incomplete due to data loss during sending or collection. If there is a lack of risk analysis of data collection and the deduplication mechanism is adjusted according to the analysis results, the integrity of data collection cannot be guaranteed.

[0003] To this end, the present invention provides a method and device for data collection of an automatic guided transport vehicle for mining. Summary of the invention

[0004] In order to make up for the deficiencies of the prior art, at least one technical problem raised in the background technology is solved.

[0005] The technical solution adopted by the present invention to solve its technical problem is:

[0006] In a first aspect, the present invention provides a method for data collection of an automatic guided transport vehicle for mining, comprising:

[0007] Step 1: In multiple historical collection cycles, similarity analysis is performed on the target data collected by the data center and the target data of the same type sent by the mining automatic guided transport vehicle. According to the analysis results, whether the target data in the historical collection cycle is collected completely is judged, and according to whether the target data collection is complete, the target data collection type in the historical collection cycle is identified. The target data collection type includes collection failure and collection success;

[0008] Step 2: Perform matrix analysis on the target data collection types within the historical collection cycle, and classify the target data into risk data and normal data based on the analysis results;

[0009] Step 3: Based on the risk data, conduct a stability analysis on the occurrence of historical collection cycles of collection failures corresponding to the risk data, and determine the predicted cycle of risk data collection failures based on the analysis results;

[0010] Step 4: Compare the predicted period of risk data collection failure with the current collection period, and determine whether deduplication optimization of risk data collection is required before the mining automatic guided transport vehicle sends risk data to the distributed message queue based on the comparison result. If necessary, generate an optimization signal;

[0011] Step 5: Based on the optimization signal, perform deduplication optimization on risk data collection.

[0012] In a second aspect, the present invention provides a device for collecting data of an automatic guided transport vehicle for mining, comprising:

[0013] Collection and analysis module: In multiple historical collection cycles, similarity analysis is performed on the target data collected by the data center and the target data of the same type sent by the mining automatic guided transport vehicle. Based on the analysis results, it is determined whether the target data in the historical collection cycle is collected completely, and the target data collection type in the historical collection cycle is identified based on whether the target data collection is complete. The target data collection type includes collection failure and collection success.

[0014] Data analysis module: performs matrix analysis on the target data collection types within the historical collection cycle, and classifies the target data into risk data and normal data based on the analysis results;

[0015] Prediction analysis module: Based on risk data, the module performs stability analysis on the occurrence of historical collection cycles of collection failures corresponding to risk data, and determines the predicted cycle of risk data collection failures based on the analysis results;

[0016] Optimization analysis module: compares the predicted period of risk data collection failure with the current collection period, and determines whether it is necessary to perform deduplication optimization of risk data collection before the mining automatic guided transport vehicle sends risk data to the distributed message queue based on the comparison result. If necessary, an optimization signal is generated;

[0017] Deduplication optimization module: Deduplication optimization is performed on risk data collection based on optimization signals.

[0018] The beneficial effects of the present invention are as follows: similarity analysis is performed on target data collected by a data center and target data of the same type sent by an automated guided transport vehicle for mining, and it is determined whether the target data within a historical collection period is collected completely. The target data collection type is identified according to the collection completeness of the target data. By performing matrix analysis on the target data collection type within a historical collection period, the target data is classified, and stability analysis is performed on the occurrence of historical collection periods of failed collection corresponding to the classified risk data. The predicted period of failed risk data collection is determined according to the analysis results, which is beneficial to the data collection of automated guided transport vehicles for mining. Predictive analysis is performed in advance to facilitate the optimization of subsequent data collection. The prediction cycle in which risk data collection fails is compared with the current collection cycle to determine whether deduplication optimization of risk data collection is required before the mine-based automated guided transport vehicle sends risk data to the distributed message queue. If necessary, the deduplication mechanism is changed in the deduplication operation before the risk data is sent, so that two copies of the risk data after deduplication are retained. This is beneficial to improving the integrity of data collection for mine-based automated guided transport vehicles in the current collection cycle, and preventing only one copy of the data from being retained for sending after deduplication. If there is a risk of loss during sending, the problem of incomplete data collection for mine-based automated guided transport vehicles may occur. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The present invention will be further described below in conjunction with the accompanying drawings.

[0020] Figure 1 It is a flowchart of the steps of a method for collecting data of an automatic guided transport vehicle for mining according to an embodiment of the present invention;

[0021] Figure 2 It is a flowchart of a data acquisition device for an automatic guided transport vehicle for mining described in an embodiment of the present invention. DETAILED DESCRIPTION

[0022] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0023] Example 1: Figure 1 As shown, a method for data collection of an automatic guided transport vehicle for mining according to an embodiment of the present invention comprises:

[0024] Step 1: In multiple historical collection cycles, similarity analysis is performed on the target data collected by the data center and the target data of the same type sent by the mining automatic guided transport vehicle. According to the analysis results, whether the target data in the historical collection cycle is collected completely is judged, and according to whether the target data collection is complete, the target data collection type in the historical collection cycle is identified. The target data collection type includes collection failure and collection success;

[0025] Specifically, the target data is obtained through the data collection report of the data collection center, wherein the target data includes but is not limited to position information, speed, load, steering angle, etc.;

[0026] The target data collected by the data center and the target data of the same type sent by the mining automated guided transport vehicle are analyzed for similarity, including:

[0027] The similarity between the target data collected by the data center and the target data of the same type sent by the mining automated guided transport vehicle is calculated using a similarity algorithm, and compared with a preset similarity;

[0028] If the similarity is greater than or equal to the preset similarity, it means that the target data collection within the historical collection period is complete;

[0029] If the similarity is less than the preset similarity, it means that the target data collection in the historical collection period is incomplete;

[0030] It should be noted that the use of a similarity algorithm to calculate the similarity between the target data collected by the data center and the target data of the same type sent by the automatic guided transport vehicle for mining includes: selecting a suitable similarity algorithm according to the type and characteristics of the target data, and the specific similarity algorithm includes but is not limited to Euclidean distance, cosine similarity, Manhattan distance and Jaccard similarity coefficient. For example, if the target data type is position information, the Euclidean distance algorithm can be used to measure the position difference of the automatic guided transport vehicle for mining in two-dimensional or three-dimensional space, such as comparing two positions or multiple points on a path; if the target data type is a steering angle, the cosine similarity algorithm can be used to measure the difference in direction of the data of the automatic guided transport vehicle for mining;

[0031] If the target data collection in the historical collection period is complete, the target data collection in the historical collection period is marked as collection success, and the historical collection period is marked as a successful collection period;

[0032] If the target data collection in the historical collection period is incomplete, the target data collection in the historical collection period is marked as collection failure, and the historical collection period is marked as a failed collection period;

[0033] Step 2: Perform matrix analysis on the target data collection types within the historical collection cycle, and classify the target data into risk data and normal data based on the analysis results;

[0034] According to the similarity between the target data collected by the data center and the target data of the same type sent by the mining automated guided transport vehicle, a similarity deviation matrix P is constructed. j ,as follows:

[0035] ;

[0036] Among them, the first column C j Indicates whether the target data collection type in the jth historical collection cycle is collection failure. 1 indicates collection failure, and 0 indicates normal collection. The second column D j It represents the similarity deviation of the target data in the jth historical collection cycle. If the collection is normal, it is 0. If the collection fails, D j =XS-XSY, where XS represents the similarity between the target data collected by the data center and the target data of the same type sent by the mining automatic guided transport vehicle when the collection type of the target data in the jth historical collection period is collection failure, and XSY represents the similarity threshold;

[0037] Based on the similarity deviation matrix P j , extract the matrix The row with value 1;

[0038] In the matrix The similarity deviation corresponding to the row with value 1 is normalized to the interval [0,1];

[0039] Exemplary: Assume that the extraction matrix After the row with 1 is obtained, the similarity deviations are 0.1, 0.2, 0.3, 0.4, 0.32, and 0.36; then the maximum value XS is extracted from the similarity deviation max and minimum value XS min , and normalize each similarity deviation: (0.1-XS min ) / (XS max- XS min ), for example, normalize 0.36: (0.36-0.1) / (0.6-0.1)=0.52;

[0040] All similarity deviations are normalized and integrated into a normalized data group, and the normalized data group is averaged to obtain the collection outliers;

[0041] comparing the acquisition anomaly value to an acquisition anomaly threshold;

[0042] If the collection abnormal value is greater than or equal to the collection abnormal threshold, the target data corresponding to the collection abnormal value is marked as risk data;

[0043] If the collection abnormal value is less than the collection abnormal threshold, the target data corresponding to the collection abnormal value is marked as normal data;

[0044] It should be noted that the collection abnormality threshold is set by those skilled in the art, and is intended to determine the collection risk of target data by the number of target data collection failures and the incompleteness (similarity deviation) of target data when collection fails;

[0045] Step 3: Based on the risk data, conduct a stability analysis on the occurrence of historical collection cycles of collection failures corresponding to the risk data, and determine the predicted cycle of risk data collection failures based on the analysis results;

[0046] Specifically, based on multiple historical collection cycles, a periodic sequence of historical collection cycles is constructed;

[0047] The moving average method is used to smooth the periodic sequence of the historical collection period, including:

[0048] Select a time window K (number of historical collection cycles);

[0049] The selection method of the number of moving windows K is as follows: the historical collection period in which the collection failed is marked as a failed collection period, and the historical collection period in which the collection succeeded is marked as a successful collection period, and the number of successful collection periods between every two adjacent failed collection periods is counted and marked as the number of successful collection period intervals, and all the successful collection period intervals are integrated into a period interval data group. If there is a mode in the period interval data group, the number of successful collection period intervals corresponding to the mode is used as the time window. If the mode is not unique, the mode is summed and the mean time window is obtained. If there is no mode in the period interval data group, the number of successful collection period intervals contained in the period interval data group is summed and the mean is obtained to obtain the time window.

[0050] According to the selected time window, the similarity mean corresponding to the historical collection period contained in each time window is calculated to obtain a smoothed period sequence;

[0051] Mark the failed acquisition cycles in the smoothed cycle sequence, obtain the number of successful acquisition cycles between every two adjacent failed acquisition cycles, and integrate them into an interval data group, calculate the variance of the interval data group, and obtain the stable value;

[0052] comparing the occurrence stability value to the occurrence stability threshold;

[0053] If the stability value is greater than the stability threshold, it means that the historical collection period of failed collection corresponding to the risk data is unstable;

[0054] If the stability value is less than or equal to the stability threshold, it means that the historical collection period of failed collection corresponding to the risk data has stabilized;

[0055] If the historical collection period of failed collection corresponding to the risk data becomes stable, the mean of the interval data group is calculated to obtain the predicted interval period;

[0056] Obtain the most recent failed collection cycle adjacent to the current collection cycle, and sum it with the prediction interval cycle to obtain the predicted cycle of risk data collection failure;

[0057] If the historical collection cycle of failed collection corresponding to the risk data is unstable, a linear regression analysis is performed on the number of successful collection cycles between adjacent failed collection cycles, and based on the analysis results, it is determined whether there is a linear change in the occurrence of failed collection cycles. Specifically:

[0058] Perform linear regression fitting on the interval data set to obtain a fitted straight line model, y=kx+m, where y represents the number of successful acquisition cycles between two adjacent failed acquisition cycles, k represents the slope, x represents the data number in the interval data set, and m represents the intercept;

[0059] Based on the interval data set and the fitted straight line model, the following variables are obtained;

[0060] The actual observation interval value yx represents the number of successful collection cycles corresponding to the xth data;

[0061] Model prediction interval value y ， x represents the predicted value of the number of successful acquisition cycles corresponding to the x-th data by the fitted straight line model;

[0062] Observation interval mean y 均 , represents the average value of all observation intervals;

[0063] Number of observations: n, which represents the total number of observations, is equal to the data number in the x-interval data group;

[0064] Calculate the total sum of squares TSS, specifically: ;

[0065] Calculate the residual sum of squares RSS, specifically: ;

[0066] Calculate the R² value as follows: ;

[0067] Compare the obtained R² value with the R² threshold;

[0068] If the R² value is greater than the R² threshold, it means that there is a linear change in the occurrence of failed acquisition cycles;

[0069] If the R² value is less than or equal to the R² threshold, it means that there is no linear change in the occurrence of failed acquisition cycles;

[0070] If the occurrence of failed collection cycles changes linearly, the prediction interval cycle is obtained according to the fitted straight line model prediction, the most recent failed collection cycle adjacent to the current collection cycle is obtained, and the sum is calculated with the prediction interval cycle to obtain the prediction cycle of risk data collection failure;

[0071] If there is no linear change in the occurrence of failed collection cycles, the mode value in the interval data group is selected as the predicted interval cycle, the most recent failed collection cycle adjacent to the current collection cycle is obtained, and the sum is calculated with the predicted interval cycle to obtain the predicted cycle of risk data collection failure;

[0072] It should be noted that if there is no mode value in the interval data group, the interval data group is divided into several interval data groups, wherein the amount of data contained in the interval data groups is equal, variance calculation is performed on the interval data groups, and the interval data group with the smallest variance value is selected for averaging to obtain the prediction interval period, wherein the prediction interval period represents the period interval between the prediction period and the most recently appeared historical collection period;

[0073] The technical solution of the embodiment of the present invention is: in multiple historical collection cycles, similarity analysis is performed on the target data collected by the data center and the same type of target data sent by the automatic guided transport vehicle for mining, to determine whether the target data in the historical collection cycle is collected completely, and to identify the target data collection type in the historical collection cycle, by performing matrix analysis on the target data collection type in the historical collection cycle, the target data is classified, and stability analysis is performed on the occurrence of historical collection cycles with failed collection corresponding to the classified risk data, and the predicted cycle of risk data collection failure is determined based on the analysis results, which is conducive to predicting and determining the cycle time of the next failed collection, and is conducive to the early prediction analysis of data collection for automatic guided transport vehicles for mining, and is convenient for the optimization of subsequent data collection.

[0074] Example 2: Figure 1 As shown, based on Example 1, a method for data collection of an automatic guided transport vehicle for mining according to an embodiment of the present invention includes:

[0075] Step 4: Compare the predicted period of risk data collection failure with the current collection period, and determine whether deduplication optimization of risk data collection is required before the mining automatic guided transport vehicle sends risk data to the distributed message queue based on the comparison result. If necessary, generate an optimization signal;

[0076] Specifically, the predicted period of risk data collection failure is compared with the current collection period;

[0077] If the prediction period of risk data collection failure coincides with the current collection period, it means that the risk of data collection failure is high in the current collection period, and risk data deduplication optimization is required, and an optimization signal is generated;

[0078] If the prediction period of risk data collection failure does not overlap with the current collection period, it means that the risk of data collection failure is small in the current collection period, and there is no need to perform deduplication optimization of risk data. In this case, no operation is performed, and the collection of mining automated guided vehicle data continues in the current collection period.

[0079] Step 5: Based on the optimization signal, perform deduplication optimization on risk data collection;

[0080] Specifically, the number of repetitions of risk data sent by the mine automatic guided transport vehicle in the current collection cycle is obtained through the data sending record report of the mine automatic guided transport vehicle data. If the number of repetitions of risk data sent by the mine automatic guided transport vehicle in the current collection cycle is equal to 2, no deduplication operation is performed before the mine automatic guided transport vehicle sends the risk data to the distributed message queue;

[0081] If the number of repetitions of risk data sent by the mining automated guided transport vehicle in the current collection cycle is greater than 2, the deduplication mechanism is changed when the mining automated guided transport vehicle performs deduplication operation before sending the risk data to the distributed message queue, so that two copies of the risk data after deduplication are retained;

[0082] The technical solution of the embodiment of the present invention is: compare the predicted period of risk data collection failure with the current collection period, and determine whether deduplication optimization of risk data collection is required before the mine-used automatic guided transport vehicle sends the risk data to the distributed message queue. If necessary, in the deduplication operation before the risk data is sent, the deduplication mechanism is changed so that two copies of the risk data after deduplication are retained, which is beneficial to improving the integrity of data collection of mine-used automatic guided transport vehicles in the current collection period, and preventing only one copy of the data from being retained for sending after deduplication, which may cause incomplete data collection of mine-used automatic guided transport vehicles due to the risk of loss during sending.

[0083] Example 3: Figure 2 As shown, a device for collecting data of an automatic guided transport vehicle for mining according to an embodiment of the present invention comprises:

[0084] Collection and analysis module: In multiple historical collection cycles, similarity analysis is performed on the target data collected by the data center and the target data of the same type sent by the mining automatic guided transport vehicle. Based on the analysis results, it is determined whether the target data in the historical collection cycle is collected completely, and the target data collection type in the historical collection cycle is identified based on whether the target data collection is complete. The target data collection type includes collection failure and collection success.

[0085] Data analysis module: performs matrix analysis on the target data collection types within the historical collection cycle, and classifies the target data into risk data and normal data based on the analysis results;

[0086] Prediction analysis module: Based on risk data, the module performs stability analysis on the occurrence of historical collection cycles of collection failures corresponding to risk data, and determines the predicted cycle of risk data collection failures based on the analysis results;

[0087] Optimization analysis module: compares the predicted period of risk data collection failure with the current collection period, and determines whether it is necessary to perform deduplication optimization of risk data collection before the mining automatic guided transport vehicle sends risk data to the distributed message queue based on the comparison result. If necessary, an optimization signal is generated;

[0088] Deduplication optimization module: Deduplication optimization is performed on risk data collection based on optimization signals.

[0089] The above shows and describes the basic principles, main features and advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for data collection of an automatic guided vehicle for mining, characterized in that: include: In multiple historical collection cycles, similarity analysis is performed on the target data collected by the data center and the target data of the same type sent by the mining automatic guided transport vehicle. Based on the analysis results, it is determined whether the target data in the historical collection cycle is collected completely, and the target data collection type in the historical collection cycle is identified based on whether the target data collection is complete. The target data collection type includes collection failure and collection success. Conduct matrix analysis on the target data collection types within the historical collection cycle, and classify the target data into risk data and normal data based on the analysis results; Based on the risk data, a stability analysis is performed on the occurrence of historical collection cycles of collection failures corresponding to the risk data, and a predicted cycle of risk data collection failure is determined based on the analysis results; Compare the predicted period of risk data collection failure with the current collection period, and determine whether deduplication optimization of risk data collection is required before the mining automatic guided transport vehicle sends risk data to the distributed message queue based on the comparison result. If necessary, generate an optimization signal; Based on the optimization signals, risk data collection is optimized for deduplication.

2. The method for data collection of a mining automatic guided vehicle according to claim 1, characterized in that: The process of judging whether the target data within the historical collection period is completely collected is as follows: The similarity between the target data collected by the data center and the target data of the same type sent by the mining automated guided transport vehicle is calculated using a similarity algorithm, and compared with a preset similarity; If the similarity is greater than or equal to the preset similarity, it means that the target data collection within the historical collection period is complete; If the similarity is less than the preset similarity, it means that the target data collection within the historical collection period is incomplete.

3. The method for data collection of an automatic guided vehicle for mining according to claim 1, characterized in that: The identification of the target data collection type is specifically as follows: If the target data collection in the historical collection period is complete, the target data collection in the historical collection period is marked as collection success, and the historical collection period is marked as a successful collection period; If the target data collection in the historical collection period is incomplete, the target data collection in the historical collection period is marked as collection failure, and the historical collection period is marked as a failed collection period.

4. The method for data collection of an automatic guided vehicle for mining according to claim 1, characterized in that: The process of classifying the target data into risk data and normal data according to the analysis results is as follows: According to the similarity between the target data collected by the data center and the target data of the same type sent by the mining automated guided transport vehicle, a similarity deviation matrix P is constructed. j ,as follows: ； Among them, the first column C j Indicates whether the target data collection type in the jth historical collection cycle is collection failure. 1 indicates collection failure, and 0 indicates normal collection. The second column D j It represents the similarity deviation of the target data in the jth historical collection period. If the collection is normal, it is 0. If the collection fails, D j =XS-XSY, where XS represents the similarity between the target data collected by the data center and the target data of the same type sent by the mining automatic guided transport vehicle when the collection type of the target data in the jth historical collection period is collection failure, and XSY represents the similarity threshold; Based on the similarity deviation matrix P j , extract the matrix The row with value 1; In the matrix The similarity deviation corresponding to the row with value 1 is normalized to the interval [0,1]; All similarity deviations are normalized and integrated into a normalized data group, and the normalized data group is averaged to obtain the collection outliers; comparing the acquisition anomaly value to an acquisition anomaly threshold; If the collection abnormal value is greater than or equal to the collection abnormal threshold, the target data corresponding to the collection abnormal value is marked as risk data; If the collection abnormal value is less than the collection abnormal threshold, the target data corresponding to the collection abnormal value is marked as normal data.

5. The method for data collection of an automatic guided vehicle for mining according to claim 1, characterized in that: The process of performing stability analysis on the occurrence of historical collection cycles of collection failures corresponding to risk data is as follows: Based on multiple historical collection cycles, a period sequence of historical collection cycles is constructed; The moving average method is used to smooth the periodic sequence of the historical collection period, including: Select a time window; According to the selected time window, the similarity mean corresponding to the historical collection period contained in each time window is calculated to obtain a smoothed period sequence; Mark the failed acquisition cycles in the smoothed cycle sequence, obtain the number of successful acquisition cycles between every two adjacent failed acquisition cycles, and integrate them into an interval data group, calculate the variance of the interval data group, and obtain the stable value; comparing the occurrence stability value to the occurrence stability threshold; If the stability value is greater than the stability threshold, it means that the historical collection period of failed collection corresponding to the risk data is unstable; If the stability value is less than or equal to the stability threshold, it means that the historical collection period of failed collection corresponding to the risk data has stabilized.

6. The method for data collection of an automatic guided vehicle for mining according to claim 5, characterized in that: The process of determining the prediction period of risk data collection failure based on the analysis results is as follows: If the historical collection period of failed collection corresponding to the risk data becomes stable, the mean of the interval data group is calculated to obtain the predicted interval period; Obtain the most recent failed collection cycle adjacent to the current collection cycle, and sum it with the prediction interval cycle to obtain the predicted cycle of risk data collection failure; If the historical collection cycle of failed collection corresponding to the risk data becomes unstable, a linear regression analysis is performed on the number of successful collection cycles between adjacent failed collection cycles, and based on the analysis results, it is determined whether there is a linear change in the occurrence of failed collection cycles. Based on whether there is a linear change, the predicted cycle of failed risk data collection is analyzed.

7. A method for data collection of an automatic guided vehicle for mining according to claim 6, characterized in that: The process of analyzing and obtaining the prediction period of risk data collection failure according to whether there is a linear change is as follows: If the occurrence of failed collection cycles changes linearly, the prediction interval cycle is obtained according to the fitted straight line model prediction, the most recent failed collection cycle adjacent to the current collection cycle is obtained, and the sum is calculated with the prediction interval cycle to obtain the prediction cycle of risk data collection failure; If there is no linear change in the occurrence of failed collection cycles, the mode value in the interval data group is selected as the predicted interval cycle, the most recent failed collection cycle adjacent to the current collection cycle is obtained, and the sum is calculated with the predicted interval cycle to obtain the predicted cycle of risk data collection failure; If there is no mode value in the interval data group, the interval data group is divided into a number of interval data groups, wherein the amount of data contained in the interval data groups is equal, variance calculation is performed on the interval data groups, and the interval data group with the smallest variance value is selected for averaging to obtain a prediction interval period, wherein the prediction interval period represents the period interval between the prediction period and the most recently occurring historical collection period; The optimization signal is generated in the following manner: Compare the predicted period of risk data collection failure with the current collection period; If the forecast period during which risk data collection fails coincides with the current collection period, an optimization signal is generated.

8. The method for data collection of an automatic guided vehicle for mining according to claim 7, characterized in that: The method for determining whether there is a linear change is as follows: Perform linear regression fitting on the interval data set, and obtain the following variables based on the interval data set and the fitted straight line model; The actual observation interval value yx represents the number of successful collection cycles corresponding to the xth data; Model prediction interval value y ， x represents the predicted value of the number of successful acquisition cycles corresponding to the x-th data by the fitted straight line model; Observation interval mean y 均 , represents the average value of all observation intervals; Number of observations: n, which represents the total number of observations, is equal to the data number in the x-interval data group; Calculate the total sum of squares TSS, specifically: ; Calculate the residual sum of squares RSS, specifically: ; Calculate the R² value as follows: ; Compare the obtained R² value with the R² threshold; If the R² value is greater than the R² threshold, it means that there is a linear change in the occurrence of failed acquisition cycles; If the R² value is less than or equal to the R² threshold, it means that there is no linear change in the occurrence of failed acquisition cycles.

9. The method for data collection of an automatic guided vehicle for mining according to claim 1, characterized in that: The process of deduplication optimization for risk data collection is as follows: The number of repetitions of risk data sent by the mine automatic guided transport vehicle in the current collection cycle is obtained through the data sending record report of the mine automatic guided transport vehicle data. If the number of repetitions of risk data sent by the mine automatic guided transport vehicle in the current collection cycle is equal to 2, no deduplication operation is performed before the mine automatic guided transport vehicle sends risk data to the distributed message queue; If the number of repetitions of risk data sent by the mining automated guided transport vehicle in the current collection cycle is greater than 2, the deduplication mechanism is changed when deduplication is performed before the mining automated guided transport vehicle sends the risk data to the distributed message queue so that two copies of the risk data after deduplication are retained.

10. A device for collecting data of an automatic guided vehicle for mining, characterized in that: include: Collection and analysis module: In multiple historical collection cycles, similarity analysis is performed on the target data collected by the data center and the target data of the same type sent by the mining automatic guided transport vehicle. Based on the analysis results, it is determined whether the target data in the historical collection cycle is collected completely, and the target data collection type in the historical collection cycle is identified based on whether the target data collection is complete. The target data collection type includes collection failure and collection success. Data analysis module: performs matrix analysis on the target data collection types within the historical collection cycle, and classifies the target data into risk data and normal data based on the analysis results; Prediction analysis module: Based on risk data, the module performs stability analysis on the occurrence of historical collection cycles of collection failures corresponding to risk data, and determines the predicted cycle of risk data collection failures based on the analysis results; Optimization analysis module: compares the predicted period of risk data collection failure with the current collection period, and determines whether it is necessary to perform deduplication optimization of risk data collection before the mining automatic guided transport vehicle sends risk data to the distributed message queue based on the comparison result. If necessary, an optimization signal is generated; Deduplication optimization module: Based on optimization signals, deduplication optimization is performed on risk data collection.

Images