An intelligent control system and method for sorting robots based on artificial intelligence

Abstract

The present invention relates to the field of intelligent sorting technology, and specifically to an intelligent control system and method for sorting robots based on artificial intelligence. The present invention verifies the signal sent by the previous robot to complete the sorting operation, and when the signal sent by the previous sorting robot is correct, the next sorting robot performs the sorting operation corresponding to the next sorting robot to avoid waste of resources caused by sorting errors; if the previous sorting robot sends an erroneous signal, the next sorting robot can remedy the unfinished sorting operation of the previous sorting robot to ensure that the sorting plan can proceed smoothly and reduce the time wasted due to downtime and waiting; in the process of verifying the signal, multi-level verification is performed, and verification is performed from multiple information sources to ensure the accuracy of the results, and the number of verification values ​​can be dynamically adjusted to achieve the purpose of dynamic verification process.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent sorting, and in particular to an intelligent management and control system and method for a sorting robot based on artificial intelligence. Background Art

[0002] Power material warehousing is the material basis for the booming development of the power industry. With the continuous increase in electricity demand, the requirements for power material warehousing are also getting higher and higher. In addition, as an important part of logistics, power material warehousing involves various links such as warehousing, inventory, outbound and return. Each link includes material category, quantity, purpose, storage location, inventory status, etc. Among them, power material warehousing is closely related to the cargo location information of power materials, and the cargo location information allocation operation of power materials is required to complete the power material warehousing.

[0003] Due to the needs of multi-station operations, multiple robots need to work together when sorting goods. Since the sorting process is time-series, each robot needs to act in sequence. When the next robot receives the signal from the previous robot, it can act, but the next robot cannot verify the accuracy of the received signal. If the previous robot sends a wrong signal, it will cause the next robot to perform an incorrect sorting operation, which will lead to waste of resources and sorting chaos. Summary of the invention

[0004] The purpose of the present invention is to provide an intelligent management and control system and method for a sorting robot based on artificial intelligence to solve the problems raised in the above-mentioned background technology.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: an intelligent management and control system for sorting robots based on artificial intelligence, the management and control system comprising a sorting module, a material planning module and a signal verification module;

[0006] The material planning module is used to determine the materials that need to be sorted, obtain the information of the materials, and formulate a sorting plan;

[0007] The sorting module includes a plurality of sorting robots, which are assigned to respective corresponding sorting operations according to the sorting plan to sort the materials; the sorting robots include the current sorting robot, the previous sorting robot and the next sorting robot;

[0008] The signal verification module is used to determine whether the signal of sorting completion sent by the current sorting robot to the next sorting robot is correct; if the received signal of sorting completion is correct, the next sorting robot sorts the materials according to the sorting plan; if the received signal of sorting completion is incorrect, the next sorting robot changes the operation of sorting;

[0009] The input end of the sorting module is electrically connected to the output ends of the material planning module and the signal verification module.

[0010] The signal verification module includes a primary verification unit, a secondary verification unit and a tertiary verification unit;

[0011] The first-level verification unit is used to randomly select the friction force data and material weight of the end effector of the current sorting robot corresponding to a certain time point in the process of the current sorting robot performing the current sorting operation, and whether the friction force data and material weight data of the end effector of the current sorting robot corresponding to a certain time point in the historical sorting event performed by the current sorting robot are the same;

[0012] The secondary verification unit is used to determine the moment when the current sorting robot completes the sorting operation and the weight change data in the corresponding sorting area, and to determine whether the weight change data is the same as the weight of the materials sorted by the current sorting robot;

[0013] The three-level verification unit is used to track the path of designated materials and determine whether they have been sorted to the designated location.

[0014] The signal verification module also includes a sorting operation learning unit, which is used to control the next sorting robot to learn the sorting operation of the current sorting robot, and perform the sorting operation of the current sorting robot on the materials. After the sorting operation of the current sorting robot is completed, the next sorting robot performs the sorting operation corresponding to the sorting plan.

[0015] The signal verification module also includes a sampling time period unit and a sampling point number unit;

[0016] The sampling time period unit is used to mark all abnormal time points on the time axis 1, extract the two abnormal time points with the longest distance on the time axis 1 as two endpoints, use the time period between the two endpoints as the warning time period, and randomly extract at least one time point in the warning time period;

[0017] The sampling point number unit is used to determine the minimum number of randomly extracted time points. If the failure rate of the current sorting robot is β<β 阈 When , the minimum number of random points is q, if the failure rate of the current sorting robot β ≥ β 阈 When , the minimum number of random points is Q.

[0018] An intelligent control method for a sorting robot based on artificial intelligence is applied to an intelligent control system for a sorting robot based on artificial intelligence. The control method includes:

[0019] Determine the materials that need to be sorted, obtain the information of the materials, formulate a sorting plan, and assign the sorting operations corresponding to each sorting robot according to the sorting plan;

[0020] After the current sorting robot receives the correct signal of the completion of the previous sorting operation sent by the previous sorting robot, the current sorting robot sorts the materials according to the sorting plan, and after the sorting is completed, it sends the signal of the completion of the current sorting operation to the next sorting robot, and the next sorting robot determines whether the received signal of the completion of the sorting operation is correct;

[0021] If the received signal indicating the completion of the sorting operation is correct, the next sorting robot will sort the materials according to the sorting plan; if the received signal indicating the completion of the sorting operation is incorrect, the next sorting robot will change the operation of the sorting operation.

[0022] The process of judging whether the received signal indicating that the sorting operation is completed is correct is as follows: performing primary verification and secondary verification on the received signal indicating that the sorting operation is completed;

[0023] The first level verification process is as follows:

[0024] Obtain the historical sorting data of the current sorting robot, and take the event of a complete sorting operation in the historical data of the current sorting robot as a sorting historical event, and the signal of the completion of the sorting operation sent by the sorting historical event is correct, and form a historical event set with multiple sorting historical events; take the sorting process of each sorting historical event as a complete timeline 1, and set the start time of all sorting historical events to T1;

[0025] Extract the complete process of the current sorting operation performed by the current sorting robot, generate timeline 2, and set the start time of timeline 2 to T2, where T1 = T2;

[0026] Randomly collect multiple time points t1, t2, ... tn on time axis 1 and time axis 2, and collect the material weight corresponding to each time point and the friction data of the actuator end of the current sorting robot, record the material weight corresponding to each time point on time axis 1 as G1, and record the friction data of the actuator end corresponding to each time point on time axis 1 as F1; record the material weight corresponding to each time point on time axis 2 as G2, and record the friction data of the actuator end corresponding to each time point on time axis 2 as F2; ​​if G1 = G2 ± α, and F1 = F2 ± α, where α is the error parameter, then it is judged that the first-level verification is passed; if G1 ≠ G2 ± α, or F1 ≠ F2 ± α, then it is judged that the first-level verification is not passed;

[0027] The process of the secondary verification is as follows: establish a corresponding relationship between the current sorting robot and the sorting area where the sorted materials are placed, obtain the time Tn when the current sorting robot determines that the sorting operation is completed, extract the weight change data in the sorting area corresponding to the time Tn, and determine whether the weight change data is the same as the weight of the materials sorted by the current sorting robot. If so, it is determined that the secondary verification has passed, otherwise, it is determined that the secondary verification has failed;

[0028] If both the first-level verification and the second-level verification are passed, it is judged that the signal of sorting completion sent by the current sorting robot is correct;

[0029] If both the first-level verification and the second-level verification fail, it is determined that the signal sent by the current sorting robot to complete the sorting operation is wrong;

[0030] If the first-level verification or the second-level verification fails, the third-level verification will be carried out;

[0031] The three-level verification process is as follows: extract the complete operation image of the current sorting operation, track the path of the specified materials, and determine whether they have been sorted to the specified location. If so, the signal that the sorting operation is completed is correct; if not, the signal that the sorting operation is completed is wrong.

[0032] The process of randomly collecting multiple time points on time axis 1 and time axis 2 is as follows:

[0033] Obtain the historical sorting data of the current sorting robot, mark the events in the historical sorting data that have not completed a complete sorting operation as sorting failure events, and obtain all sorting failure events that appear in the historical sorting data;

[0034] According to the number of occurrences of sorting failure events m, the sorting failure rate β of the current sorting robot is calculated. The calculation formula is β=m / M, where M represents all sorting events of the current sorting robot; if the failure rate β of the current sorting robot is less than β 阈 When , the minimum number of random points is q, if the failure rate of the current sorting robot β ≥ β 阈 When , the minimum number of random points is Q, and Q>q; where β 阈 is the set failure rate threshold;

[0035] The formula for calculating Q is as follows: Q=β*k+q, where k is a preset value and k is a positive number, and * represents a multiplication sign;

[0036] Extract the sorting failure cause corresponding to each sorting failure event, mark the data used to determine the sorting failure cause as abnormal data, mark the time point when the abnormal data appears as the abnormal time point, mark all abnormal time points on time axis one, extract the two abnormal time points with the longest distance on time axis one as the two endpoints, take the time period between the two endpoints as the warning time period, and randomly extract at least one time point within the warning time period.

[0037] The process of the next sorting robot changing the operation of the sorting operation is as follows: if the signal received by the next sorting robot that the sorting operation is completed is wrong, the sorting instruction data of the current sorting robot is packaged and sent to the controller of the next sorting robot, the next sorting robot learns the sorting operation of the current sorting robot, and performs the sorting operation of the current sorting robot on the materials. After the sorting operation of the current sorting robot is completed, the next sorting robot performs the sorting operation corresponding to the sorting plan.

[0038] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0039] 1. The present invention verifies the signal of the completion of the sorting operation sent by the previous robot. When the signal sent by the previous sorting robot is correct, the next sorting robot performs the corresponding sorting operation to avoid waste of resources caused by sorting errors. If the previous sorting robot sends an erroneous signal, the next sorting robot can remedy the unfinished sorting operation of the previous sorting robot to ensure that the sorting plan can be carried out smoothly and reduce the time wasted due to downtime and waiting. In the process of verifying the signal, multi-level verification is carried out, and verification is carried out from multiple information sources to ensure the accuracy of the result, and the number of verification values ​​can be dynamically adjusted to achieve the purpose of dynamic verification process. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0041] Figure 1 It is a flow chart of an intelligent control method of a sorting robot based on artificial intelligence of the present invention. DETAILED DESCRIPTION

[0042] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] Embodiment 1, the present invention provides a technical solution: an intelligent management and control system of a sorting robot based on artificial intelligence, the management and control system comprising a sorting module, a material planning module and a signal verification module;

[0044] The material planning module is used to determine the materials that need to be sorted, obtain the information of the materials, and formulate a sorting plan;

[0045] The sorting module includes a plurality of sorting robots, which are assigned to respective corresponding sorting operations according to the sorting plan to sort the materials; the sorting robots include the current sorting robot, the previous sorting robot and the next sorting robot;

[0046] The signal verification module is used to determine whether the signal of sorting completion sent by the current sorting robot to the next sorting robot is correct; if the received signal of sorting completion is correct, the next sorting robot sorts the materials according to the sorting plan; if the received signal of sorting completion is incorrect, the next sorting robot changes the operation of sorting;

[0047] The input end of the sorting module is electrically connected to the output ends of the material planning module and the signal verification module.

[0048] The signal verification module includes a primary verification unit, a secondary verification unit and a tertiary verification unit;

[0049] The first-level verification unit is used to randomly select the friction force data and material weight of the end effector of the current sorting robot corresponding to a certain time point in the process of the current sorting robot performing the current sorting operation, and whether the friction force data and material weight data of the end effector of the current sorting robot corresponding to a certain time point in the historical sorting event performed by the current sorting robot are the same;

[0050] The secondary verification unit is used to determine the moment when the current sorting robot completes the sorting operation and the weight change data in the corresponding sorting area, and to determine whether the weight change data is the same as the weight of the materials sorted by the current sorting robot;

[0051] The three-level verification unit is used to track the path of designated materials and determine whether they have been sorted to the designated location.

[0052] The signal verification module also includes a sorting operation learning unit, which is used to control the next sorting robot to learn the sorting operation of the current sorting robot, and perform the sorting operation of the current sorting robot on the materials. After the sorting operation of the current sorting robot is completed, the next sorting robot performs the sorting operation corresponding to the sorting plan.

[0053] The signal verification module also includes a sampling time period unit and a sampling point number unit;

[0054] The sampling time period unit is used to mark all abnormal time points on the time axis 1, extract the two abnormal time points with the longest distance on the time axis 1 as two endpoints, use the time period between the two endpoints as the warning time period, and randomly extract at least one time point in the warning time period;

[0055] The sampling point number unit is used to determine the minimum number of randomly extracted time points. If the failure rate of the current sorting robot is β<β 阈 When , the minimum number of random points is q, if the failure rate of the current sorting robot β ≥ β 阈 When , the minimum number of random points is Q.

[0056] For example 2, please refer to Figure 1 , an intelligent control method of a sorting robot based on artificial intelligence, which is applied to an intelligent control system of a sorting robot based on artificial intelligence, the control method comprising:

[0057] Determine the materials that need to be sorted, obtain the information of the materials, formulate a sorting plan, and assign the sorting operations corresponding to each sorting robot according to the sorting plan;

[0058] After the current sorting robot receives the correct signal of the completion of the previous sorting operation sent by the previous sorting robot, the current sorting robot sorts the materials according to the sorting plan, and after the sorting is completed, it sends the signal of the completion of the current sorting operation to the next sorting robot, and the next sorting robot determines whether the received signal of the completion of the sorting operation is correct;

[0059] If the received signal indicating the completion of the sorting operation is correct, the next sorting robot will sort the materials according to the sorting plan; if the received signal indicating the completion of the sorting operation is incorrect, the next sorting robot will change the operation of the sorting operation.

[0060] The process of judging whether the received signal indicating that the sorting operation is completed is correct is as follows: performing primary verification and secondary verification on the received signal indicating that the sorting operation is completed;

[0061] The first level verification process is as follows:

[0062] Obtain the historical sorting data of the current sorting robot, and take the event of a complete sorting operation in the historical data of the current sorting robot as a sorting historical event, and the signal of the completion of the sorting operation sent by the sorting historical event is correct, and form a historical event set with multiple sorting historical events; take the sorting process of each sorting historical event as a complete timeline 1, and set the start time of all sorting historical events to T1;

[0063] Extract the complete process of the current sorting operation performed by the current sorting robot, generate timeline 2, and set the start time of timeline 2 to T2, where T1 = T2;

[0064] Randomly collect multiple time points t1, t2, ... tn on time axis 1 and time axis 2, and collect the material weight corresponding to each time point and the friction data of the actuator end of the current sorting robot, record the material weight corresponding to each time point on time axis 1 as G1, and record the friction data of the actuator end corresponding to each time point on time axis 1 as F1; record the material weight corresponding to each time point on time axis 2 as G2, and record the friction data of the actuator end corresponding to each time point on time axis 2 as F2; ​​if G1 = G2 ± α, and F1 = F2 ± α, where α is the error parameter, then it is judged that the first-level verification is passed; if G1 ≠ G2 ± α, or F1 ≠ F2 ± α, then it is judged that the first-level verification is not passed;

[0065] For example, the process of sorting historical events is formed into a timeline 1, and the process of the current sorting operation performed by the current sorting robot is formed into a timeline 2. The start time of both timeline 1 and timeline 2 is set to 00:00. According to the operation sequence of the sorting operation, a time point can be extracted on timeline 1, such as 01:00, and 01:00 can also be extracted on timeline 2. The material weight corresponding to these two points and the friction force data of the actuator end of the current sorting robot are obtained to determine whether the data at these two points are the same. If they are the same, the first-level verification is passed, and if they are not the same, the first-level verification is failed.

[0066] The process of the secondary verification is as follows: establish a corresponding relationship between the current sorting robot and the sorting area where the sorted materials are placed, obtain the time Tn when the current sorting robot determines that the sorting operation is completed, extract the weight change data in the sorting area corresponding to the time Tn, and determine whether the weight change data is the same as the weight of the materials sorted by the current sorting robot. If so, it is determined that the secondary verification has passed, otherwise, it is determined that the secondary verification has failed;

[0067] For example, the current sorting robot is sorting materials weighing 0.5kg, and the time of completing the sorting operation is 03:00. At this time, the weight change in the sorting area at that time is extracted to see whether it is 0.5kg. If so, the secondary verification is passed, if not, the secondary verification is failed.

[0068] If both the first-level verification and the second-level verification are passed, it is judged that the signal of sorting completion sent by the current sorting robot is correct;

[0069] If both the first-level verification and the second-level verification fail, it is determined that the signal sent by the current sorting robot to complete the sorting operation is wrong;

[0070] If the first-level verification or the second-level verification fails, the third-level verification will be carried out;

[0071] The three-level verification process is as follows: extract the complete operation image of the current sorting operation, track the path of the specified materials, and determine whether they have been sorted to the specified location. If so, the signal that the sorting operation is completed is correct; if not, the signal that the sorting operation is completed is wrong.

[0072] For example, manual judgment can be made by manually retrieving the operation image to determine whether the sorting operation is completed, or an image algorithm can be used to track the path of the materials, mark the designated materials as tracking points, mark the designated location as the end point, and determine whether the tracking point eventually coincides with the end point. If so, the signal indicating the completion of the sorting operation is correct; if not, the signal indicating the completion of the sorting operation is incorrect.

[0073] The process of randomly collecting multiple time points on time axis 1 and time axis 2 is as follows:

[0074] Obtain the historical sorting data of the current sorting robot, mark the events in the historical sorting data that have not completed a complete sorting operation as sorting failure events, and obtain all sorting failure events that appear in the historical sorting data;

[0075] According to the number of occurrences of sorting failure events m, the sorting failure rate β of the current sorting robot is calculated. The calculation formula is β=m / M, where M represents all sorting events of the current sorting robot; if the failure rate β of the current sorting robot is less than β 阈 When , the minimum number of random points is q, if the failure rate of the current sorting robot β ≥ β 阈 When , the minimum number of random points is Q, and Q>q; where β 阈 is the set failure rate threshold;

[0076] The formula for calculating Q is as follows: Q=β*k+q, where k is a preset value and is a positive number, and * represents a multiplication sign; Q is rounded to an integer.

[0077] Extract the sorting failure cause corresponding to each sorting failure event, mark the data used to determine the sorting failure cause as abnormal data, mark the time point when the abnormal data appears as the abnormal time point, mark all abnormal time points on time axis one, extract the two abnormal time points with the longest distance on time axis one as the two endpoints, take the time period between the two endpoints as the warning time period, and randomly extract at least one time point within the warning time period.

[0078] For example, a sorting failure event process generates a timeline 1, extracts the sorting failure cause, such as material falling off, and obtains abnormal data for judging material falling off, such as the friction data at the end of the actuator of the sorting robot changes from 10 to 0, indicating that the material falls from the end of the actuator. Then the friction data is used as abnormal data, and the time point when the friction data changes is used as the abnormal time point, which is marked on the timeline 1. The abnormal time points at both ends are extracted to form a warning time period. At least one time point is randomly selected in the warning time period, and the total number of randomly selected time points is not less than q=3. When the failure rate increases, the number of randomly selected time points will also increase accordingly. For example, Q=0.1*10+3=4, then when the failure rate is 0.1, 4 time points need to be randomly selected.

[0079] The process of the next sorting robot changing the operation of the sorting operation is as follows: if the signal received by the next sorting robot that the sorting operation is completed is wrong, the sorting instruction data of the current sorting robot is packaged and sent to the controller of the next sorting robot, the next sorting robot learns the sorting operation of the current sorting robot, and performs the sorting operation of the current sorting robot on the materials. After the sorting operation of the current sorting robot is completed, the next sorting robot performs the sorting operation corresponding to the sorting plan.

[0080] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An intelligent control system for sorting robots based on artificial intelligence, characterized by: The management and control system includes a sorting module, a material planning module and a signal verification module; The material planning module is used to determine the materials that need to be sorted, obtain the information of the materials, and formulate a sorting plan; The sorting module includes a plurality of sorting robots, which are assigned to respective corresponding sorting operations according to the sorting plan to sort the materials; the sorting robots include the current sorting robot, the previous sorting robot and the next sorting robot; The signal verification module is used to determine whether the signal of sorting completion sent by the current sorting robot to the next sorting robot is correct; if the received signal of sorting completion is correct, the next sorting robot sorts the materials according to the sorting plan; if the received signal of sorting completion is incorrect, the next sorting robot changes the operation of sorting; The input end of the sorting module is electrically connected to the output ends of the material planning module and the signal verification module; The signal verification module includes a primary verification unit, a secondary verification unit and a tertiary verification unit; The first-level verification unit is used to randomly select the friction force data and material weight of the end effector of the current sorting robot corresponding to a certain time point in the process of the current sorting robot performing the current sorting operation, and whether the friction force data and material weight data of the end effector of the current sorting robot corresponding to a certain time point in the historical sorting event performed by the current sorting robot are the same; The secondary verification unit is used to determine the moment when the current sorting robot completes the sorting operation and the weight change data in the corresponding sorting area, and to determine whether the weight change data is the same as the weight of the materials sorted by the current sorting robot; The third-level verification unit is used to track the path of the specified materials and determine whether they have been sorted to the specified location; The signal verification module also includes a sampling time period unit and a sampling number unit. The sampling time period unit is used to mark all abnormal time points on the time axis 1, extract the two abnormal time points with the longest distance on the time axis 1 as two endpoints, use the time period between the two endpoints as the warning time period, and randomly extract at least one time point in the warning time period; The sampling point number unit is used to determine the minimum number of randomly extracted time points. If the failure rate β of the current sorting robot is less than the β threshold, the minimum number of randomly selected points is q. If the failure rate β of the current sorting robot is greater than or equal to the β threshold, the minimum number of randomly selected points is calculated to be Q.

2. According to claim 1, an intelligent management and control system for sorting robots based on artificial intelligence is characterized by: The signal verification module also includes a sorting operation learning unit, which is used to control the next sorting robot to learn the sorting operation of the current sorting robot, and perform the sorting operation of the current sorting robot on the materials. After the sorting operation of the current sorting robot is completed, the next sorting robot performs the sorting operation corresponding to the sorting plan.

3. An intelligent control method for a sorting robot based on artificial intelligence, which is applied to an intelligent control system for a sorting robot based on artificial intelligence as described in any one of claims 1-2, characterized in that: The control method includes: Determine the materials that need to be sorted, obtain the information of the materials, formulate a sorting plan, and assign the sorting operations corresponding to each sorting robot according to the sorting plan; After the current sorting robot receives the correct signal of the completion of the previous sorting operation sent by the previous sorting robot, the current sorting robot sorts the materials according to the sorting plan, and after the sorting is completed, it sends the signal of the completion of the current sorting operation to the next sorting robot, and the next sorting robot determines whether the received signal of the completion of the sorting operation is correct; If the received signal indicating the completion of the sorting operation is correct, the next sorting robot will sort the materials according to the sorting plan; if the received signal indicating the completion of the sorting operation is incorrect, the next sorting robot will change the operation of the sorting operation.

4. The method for intelligent control of a sorting robot based on artificial intelligence according to claim 3, characterized in that: The process of judging whether the received signal indicating that the sorting operation is completed is correct is as follows: performing primary verification and secondary verification on the received signal indicating that the sorting operation is completed; The first level verification process is as follows: Obtain the historical sorting data of the current sorting robot, and take the event of a complete sorting operation in the historical data of the current sorting robot as a sorting historical event, and the signal of the completion of the sorting operation sent by the sorting historical event is correct, and form a historical event set with multiple sorting historical events; take the sorting process of each sorting historical event as a complete timeline 1, and set the start time of all sorting historical events to T1; Extract the complete process of the current sorting operation performed by the current sorting robot, generate timeline 2, and set the start time of timeline 2 to T2; Randomly collect multiple time points t1, t2, ... tn on time axis 1 and time axis 2, and collect the material weight corresponding to each time point and the friction data of the actuator end of the current sorting robot, record the material weight corresponding to each time point on time axis 1 as G1, and record the friction data of the actuator end corresponding to each time point on time axis 1 as F1; record the material weight corresponding to each time point on time axis 2 as G2, and record the friction data of the actuator end corresponding to each time point on time axis 2 as F2; ​​if G1 = G2 ± α, and F1 = F2 ± α, where α is the error parameter, then it is judged that the first-level verification is passed; if G1 ≠ G2 ± α, or F1 ≠ F2 ± α, then it is judged that the first-level verification is not passed; The process of the secondary verification is as follows: establish a corresponding relationship between the current sorting robot and the sorting area where the sorted materials are placed, obtain the time Tn when the current sorting robot determines that the sorting operation is completed, extract the weight change data in the sorting area corresponding to the time Tn, and determine whether the weight change data is the same as the weight of the materials sorted by the current sorting robot. If so, it is determined that the secondary verification has passed, otherwise, it is determined that the secondary verification has failed; If both the first-level verification and the second-level verification are passed, it is judged that the signal of sorting completion sent by the current sorting robot is correct; If both the first-level verification and the second-level verification fail, it is determined that the signal sent by the current sorting robot to complete the sorting operation is wrong; If the first-level verification or the second-level verification fails, the third-level verification will be carried out; The three-level verification process is as follows: extract the complete operation image of the current sorting operation, track the path of the specified materials, and determine whether they have been sorted to the specified location. If so, the signal that the sorting operation is completed is correct; if not, the signal that the sorting operation is completed is wrong.

5. The method for intelligent control of a sorting robot based on artificial intelligence according to claim 3, characterized in that: The process of randomly collecting multiple time points on time axis 1 and time axis 2 is as follows: Obtain the historical sorting data of the current sorting robot, mark the events in the historical sorting data that have not completed a complete sorting operation as sorting failure events, and obtain all sorting failure events that appear in the historical sorting data; According to the number of occurrences of sorting failure events m, the sorting failure rate β of the current sorting robot is calculated, and the calculation formula is β = m / M, where M represents all sorting events of the current sorting robot; if the failure rate β of the current sorting robot is less than β 阈 When , the minimum number of random points is q, if the failure rate of the current sorting robot β ≥ β 阈 When , the minimum number of random points is Q, and Q>q; where β 阈 is the set failure rate threshold; The formula for calculating Q is as follows: Q = β*k + q, where k is a preset value, k is a positive number, and * represents a multiplication sign; Extract the sorting failure cause corresponding to each sorting failure event, mark the data used to determine the sorting failure cause as abnormal data, mark the time point when the abnormal data appears as the abnormal time point, mark all abnormal time points on time axis one, extract the two abnormal time points with the longest distance on time axis one as the two endpoints, take the time period between the two endpoints as the warning time period, and randomly extract at least one time point within the warning time period.

6. The method for intelligent control of a sorting robot based on artificial intelligence according to claim 3, characterized in that: The process of the next sorting robot changing the operation of the sorting operation is as follows: if the signal received by the next sorting robot that the sorting operation is completed is wrong, the sorting instruction data of the current sorting robot is packaged and sent to the controller of the next sorting robot, the next sorting robot learns the sorting operation of the current sorting robot, and performs the sorting operation of the current sorting robot on the materials. After the sorting operation of the current sorting robot is completed, the next sorting robot performs the sorting operation corresponding to the sorting plan.

Images