Robot state data preprocessing method, device and equipment

Abstract

The present disclosure proposes a method, device and equipment for preprocessing robot status data, the method comprising: reading the index value of each data indicator from the status data currently reported by the robot; determining the preprocessing method corresponding to each data indicator according to the fault data of the robot; preprocessing the index value of each data indicator according to the preprocessing method corresponding to each data indicator, so that the preprocessing method corresponding to each data indicator can be dynamically adjusted according to the fault change of the robot, so that the preprocessed status data meets the needs of dynamic fault diagnosis and analysis of the robot.

Description

Technical Field

[0001] The present disclosure relates to the technical field of data processing, and in particular to a method, device and equipment for preprocessing robot state data. Background Art

[0002] With the development and application of the combination of industrial robots and the Internet, fault diagnosis and data analysis of industrial robots have been widely used with the popularization of the Internet. In the intelligent monitoring system of industrial robots, industrial robots are combined with the Internet of Things (IoT) to collect the status data of each mechanism of the robot, and perform fault diagnosis on the robot based on the status data of each mechanism. However, in the process of collecting, reporting, transferring and other processing of the status data of each mechanism, invalid data such as noise data, inconsistent data and missing data will appear, which will affect the accuracy of fault diagnosis. Therefore, how to pre-process the status data of each mechanism is particularly important.

[0003] At present, when preprocessing the status data of each mechanism, invalid data is usually deleted directly or filled with data collected "nearby". However, industrial robots are application-oriented devices. As the life cycle of their bodies changes, the factors that affect failures also change, and the requirements for status data for fault diagnosis and analysis are also different. Therefore, the status data obtained by the existing fixed preprocessing method is difficult to meet the needs of dynamic fault diagnosis and analysis. Summary of the invention

[0004] In view of this, the present disclosure proposes a method, device and equipment for preprocessing robot state data. The state data obtained by the preprocessing method of the present disclosure can meet the needs of dynamic fault diagnosis and analysis of the robot.

[0005] According to a first aspect of the present disclosure, a method for preprocessing robot state data is provided, comprising:

[0006] Read the indicator value of each data indicator from the status data currently reported by the robot;

[0007] Determine a preprocessing method corresponding to each of the data indicators according to the fault data of the robot;

[0008] The indicator value of each of the data indicators is preprocessed according to the preprocessing method corresponding to each of the data indicators.

[0009] In a possible implementation, when determining the preprocessing method corresponding to each of the data indicators according to the fault data of the robot, it includes:

[0010] Calculating the degree of fault impact corresponding to each of the data indicators according to the fault data of the robot;

[0011] According to the fault impact degree corresponding to each of the data indicators, a preprocessing method corresponding to each of the data indicators is determined.

[0012] In a possible implementation, the fault data includes at least one of a dynamic fault proportion and a preset fault proportion corresponding to each of the data indicators;

[0013] The dynamic fault proportion corresponding to each of the data indicators is calculated based on the fault frequency and fault history data corresponding to each of the data indicators.

[0014] In a possible implementation, the calculation formula for the fault impact degree is as follows:

[0015]

[0016] Where f(x) is the fault impact degree corresponding to the data indicator x, H(x i ) is the historical indicator value of the data indicator x in the i-th historical status data, d is the dynamic fault ratio corresponding to the data indicator x, p is the preset fault ratio, and T(j) is the data indicator threshold corresponding to each mechanism of the robot.

[0017] In a possible implementation, determining a preprocessing method corresponding to each of the data indicators according to the degree of impact of the fault corresponding to the data indicator includes:

[0018] According to the degree of failure impact corresponding to each of the data indicators, each of the data indicators is divided into a critical data indicator and a non-critical data indicator;

[0019] The preset first preprocessing method is used as the preprocessing method corresponding to the key data indicator, and the preset second preprocessing method is used as the preprocessing method for the non-key data indicator.

[0020] In a possible implementation, the first preprocessing method includes filling the invalid indicator value of the data indicator using a preset filling algorithm;

[0021] When the invalid indicator value of the data indicator is filled using a preset filling algorithm, it includes:

[0022] Determining whether the indicator value of the data indicator is an invalid indicator value;

[0023] When it is determined that the indicator value of the data indicator is an invalid indicator value, the invalid indicator value is revised based on a preset fault proportion corresponding to the data indicator.

[0024] In a possible implementation, when revising the invalid indicator value based on the preset fault proportion corresponding to the data indicator, it includes:

[0025] Determine whether the preset fault ratio is greater than a preset fault ratio threshold;

[0026] When it is determined that the preset fault proportion is greater than the fault proportion threshold, revising the invalid indicator value based on the indicator value of the data indicator reported before the fault;

[0027] When it is determined that the preset fault proportion is less than or equal to the fault proportion threshold, the invalid data indicator value is revised based on the indicator value of the data indicator reported in the previous item.

[0028] In a possible implementation, the second preprocessing method includes deleting invalid indicator values ​​of the data indicators.

[0029] According to a second aspect of the present disclosure, a robot state data preprocessing device is provided, comprising:

[0030] The data acquisition module is used to read the indicator value of each data indicator from the status data currently reported by the robot;

[0031] A processing method determination module, used to determine a preprocessing method corresponding to each of the data indicators according to the fault data of the robot;

[0032] The preprocessing module is used to preprocess the indicator value of each data indicator according to the preprocessing method corresponding to each data indicator.

[0033] According to a third aspect of the present disclosure, a robot state data preprocessing device is provided, comprising: a processor; a memory for storing processor executable instructions; wherein the processor is configured to execute the method described in the first aspect of the present disclosure.

[0034] When preprocessing the status data of the robot, the present invention first determines the preprocessing method corresponding to each data indicator in the status data according to the fault data of the robot, and then preprocesses it according to the preprocessing method corresponding to each data indicator. In this way, the preprocessing method corresponding to each data indicator can be dynamically adjusted according to the fault change of the robot, so that the preprocessed status data meets the needs of dynamic fault diagnosis and analysis of the robot.

[0035] Further features and aspects of the present disclosure will become apparent from the following detailed description of exemplary embodiments with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate exemplary embodiments, features, and aspects of the disclosure and, together with the description, serve to explain the principles of the disclosure.

[0037] Figure 1 A schematic flow chart showing a method for preprocessing robot state data according to an embodiment of the present disclosure;

[0038] Figure 2 A schematic flow chart showing an example of a method for preprocessing robot state data according to an embodiment of the present disclosure;

[0039] Figure 3 A schematic block diagram showing a device for preprocessing robot state data according to an embodiment of the present disclosure;

[0040] Figure 4 A schematic block diagram of a robot state data preprocessing device according to an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0041] Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0042] The word “exemplary” is used exclusively herein to mean “serving as an example, example, or illustration.” Any embodiment described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments.

[0043] In addition, in order to better illustrate the present disclosure, numerous specific details are given in the following specific embodiments. It should be understood by those skilled in the art that the present disclosure can also be implemented without certain specific details. In some examples, methods, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the subject matter of the present disclosure.

[0044] <Method Example>

[0045] Figure 1 FIG. 1 is a schematic flow chart showing a method for preprocessing robot state data according to an embodiment of the present disclosure. Figure 1 As shown, the method includes steps S1100-S1300.

[0046] S1100, reading the indicator value of each data indicator from the status data currently reported by the robot.

[0047] During operation, the robot will collect data indicators of each mechanism according to a preset clock frequency (such as 500HZ), and report the data indicators of each mechanism collected at the same clock frequency to the cloud platform as a state data (that is, a state data includes data indicators corresponding to each mechanism). Among them, the various mechanisms constituting the robot body include at least one of an actuator, a driving mechanism, a control mechanism and a sensing mechanism.

[0048] For the actuator, data indicators reflecting its operating status may include: current, voltage, temperature, humidity, and the operating status of the current hardware in the actuator - the agreed protocol values ​​are: 1 (working), 2 (not working), 3 (faulty) and at least one of the execution speed.

[0049] The driving mode of the driving mechanism of the robot may be electric drive, liquid drive, gas drive, or a hybrid drive of electric, liquid and gas, which is not specifically limited here. For an electric drive mechanism, the data indicator reflecting its operating state may include: at least one of current, voltage and temperature. For a liquid drive mechanism, the data indicator reflecting its operating state may include: at least one of hydraulic pressure, humidity and temperature. For a gas drive mechanism, the data indicator reflecting its operating state may include: at least one of pressure, temperature and humidity. For a hybrid drive mechanism, all data indicators corresponding to different driving modes may be used as data indicators reflecting its operating state, or some data indicators may be selected from all data indicators corresponding to different driving modes as data indicators reflecting its operating state, which is not specifically limited here.

[0050] For the control mechanism, data indicators reflecting its operating status may include: the rotational speed of the gear, the torque of the gear, the angular velocity of the gear, the speed of the driving conveyor belt, the transmission speed, the temperature of the component, the state of the motor, and at least one of the electrical energy of the motor and the kinetic energy of the motor.

[0051] For the sensing mechanism, the data indicators reflecting its operating status may include: at least one of temperature, humidity, current, and voltage.

[0052] S1200, determining a preprocessing method corresponding to each data indicator according to the robot's fault data.

[0053] In a possible implementation, when determining the preprocessing method corresponding to each data indicator according to the fault data of the robot, steps S1210-S1220 are included.

[0054] S1210, calculating the degree of fault impact corresponding to each data indicator according to the fault data of the robot.

[0055] Each time a robot fails, a corresponding fault data will be generated. Each fault data record at least the machine status data (i.e., fault history data) under the fault, the dynamic fault ratio corresponding to each data indicator, and the preset fault ratio. The dynamic fault ratio corresponding to each data indicator is calculated based on the fault frequency and fault history data corresponding to each data indicator.

[0056] In a possible implementation, the calculation formula for the dynamic fault ratio includes formula (1) and formula (2):

[0057] d=k / K(x) (1)

[0058]

[0059] In the formula, d is the dynamic fault proportion corresponding to the data indicator x, k is the first preset proportional coefficient, the value range of the first preset proportional coefficient k can be 1-5, K(x) is the fault dynamic impact factor corresponding to the data indicator x, m is the second preset proportional coefficient, the value range of the second preset proportional coefficient m can be 1-5, F is the fault frequency corresponding to the data indicator x, and R(x) is the indicator value of the data indicator x in the fault history data.

[0060] It should be noted that the fault frequency corresponding to each data indicator is equal to the number of faults caused by the abnormality of the data indicator divided by the total number of faults. For example, the number of faults caused by the abnormal current of the actuator is 5, and the total number of faults is 20, then the fault frequency corresponding to the data indicator of the actuator current is 5 / 20=0.2.

[0061] It should be noted that in formula (2), ∑R(x) is the sum of the index values ​​of the data index x in each fault history data. For example, if the number of faults is 10, the corresponding fault history data includes 10 pieces. In this case, when calculating ∑R(x), the index values ​​of the data index x in the 10 fault history data need to be summed.

[0062] It should be noted that after a robot fails, the engineer will locate and resolve the fault, and after resolving the fault, the engineer will set the proportion of faults caused by abnormal data indicators to the total faults based on their own fault handling experience. This proportion is the preset fault ratio. For example, after completing the fault handling, the engineer determines that the fault is caused by abnormal current of the actuator, and based on their own fault handling experience, they believe that the faults caused by abnormal current of the actuator account for 5% of the total faults. In this case, the preset fault ratio corresponding to the data indicator of actuator current can be set to 5%.

[0063] In an embodiment where the fault data at least includes a dynamic fault ratio and a preset fault ratio corresponding to each data indicator, when calculating the fault impact degree corresponding to each data indicator based on the fault data of the robot, it can be achieved by using the calculation formula (3) for the fault impact degree.

[0064]

[0065] Where f(x) is the fault impact degree corresponding to the data indicator x, H(x i ) is the index value of data index x in the i-th historical status data, n is the number of historical status data, d is the dynamic fault proportion corresponding to data index x, p is the preset fault proportion, and T(j) is the data index threshold corresponding to organization j.

[0066] It should be noted that each data indicator corresponding to each mechanism has a corresponding indicator threshold value, and the indicator threshold value corresponding to each data indicator of the same mechanism is added together to obtain the data indicator threshold value corresponding to the mechanism. Among them, the indicator threshold value is the threshold value for the normal operation of each mechanism. For example, the data indicators of the sensor mechanism include current, voltage, temperature and humidity. The rated voltage of the sensor mechanism is 220v. When the indicator value of the voltage exceeds 220v, it means that the sensor mechanism may have a short circuit or abnormality. When the indicator value of the voltage is lower than 220v, the sensor mechanism will cause the deviation of the collected data to be too large due to insufficient power. Therefore, the rated voltage 220v can be used as the indicator threshold value of the voltage. Similarly, the rated current 30A can be used as the indicator threshold value of the current, the maximum temperature value of the normal operation of the sensor mechanism 40 can be used as the indicator threshold value of the temperature, the maximum humidity value of the normal operation of the sensor mechanism is 80%, and 80 can be used as the indicator threshold value of the humidity. Therefore, for the sensor mechanism 4, its corresponding data indicator threshold value T(4)=220+30+40+80=370, that is, the data indicator threshold value corresponding to the sensor mechanism is 370. Referring to the method for determining the threshold of the sensing mechanism, the numerical index threshold T(1) corresponding to the actuator 1, the numerical index threshold T(2) corresponding to the driving mechanism 2, and the numerical index threshold T(3) corresponding to the control mechanism 3 can be determined, which will not be repeated here. After obtaining the data index threshold corresponding to each mechanism, the sum can be obtained to obtain ∑T(j) in formula (3), that is, ∑T(j)=T(1)+T(2)+T(3)+T(4).

[0067] It should be noted that before the currently reported status data, the status data reported by the robot under normal operating status is processed in sequence including preprocessing, data cleaning, data integration and data conversion to obtain regular status data, which are then stored in the database. These regular status data are the historical status data.

[0068] S1220, determining a preprocessing method corresponding to each data indicator according to the fault impact degree corresponding to each data indicator. The specific steps include S1221-S1222.

[0069] S1221, dividing each data indicator into a critical data indicator and a non-critical data indicator according to the fault impact degree corresponding to each data indicator.

[0070] In a possible implementation, after obtaining the fault impact degree corresponding to each data indicator, the data indicators are sorted according to the magnitude of the fault impact degree, and the data indicator of the set data with the largest fault impact degree is used as the key data indicator, and the remaining indicators are used as non-key indicators. Among them, the set data can be 3, that is, the 3 data indicators with the largest fault impact degree are used as key data indicators, and the remaining indicators are used as non-key indicators.

[0071] S1222: Use the preset first preprocessing method as the preprocessing method corresponding to the key data indicator, and use the preset second preprocessing method as the preprocessing method for the non-key data indicator.

[0072] In a possible implementation, the first preprocessing method includes using a preset filling algorithm to fill in the invalid indicator value of the data indicator. When the preset filling algorithm is used to fill in the invalid indicator value of the data indicator, steps S1222-1 to S1222-2 are included.

[0073] S1222-1, determine whether the indicator value of the data indicator is an invalid indicator value.

[0074] It should be noted that when the indicator value of a data indicator is lost, inconsistent with the actual indicator value, or seriously deviates from the actual indicator value, it is an invalid indicator value.

[0075] When judging whether the index value of a data index is an invalid index value, first judge whether the index value of the data index is lost. If the index value is lost, judge the index value of the data index to be an invalid index value. If the index value is not lost, continue to judge whether the currently acquired index value is inconsistent with the actual index value or seriously deviates from the actual index value. The specific steps include: judging whether the variance of the currently acquired index value and the average index value of the data index in the historical status data is greater than or equal to the set variance threshold. When it is greater than or equal to the set variance threshold, it means that the currently acquired index value is inconsistent with the actual index value or seriously deviates from the actual index value. At this time, judge the index value corresponding to the data index to be an invalid index value. Among them, the set variance threshold can be set to 0.5.

[0076] When the indicator value of the data indicator is not empty and the variance between the indicator value and the average indicator value of the data indicator in the historical status data is less than the set variance threshold, it means that the indicator of the data indicator is valid and no preprocessing operation is required.

[0077] S1222-2, when it is determined that the indicator value of the data indicator is an invalid indicator value, the invalid indicator value is revised based on the preset fault ratio corresponding to the data indicator. The specific steps include: determining whether the preset fault ratio corresponding to the data indicator is greater than the preset fault ratio threshold; when it is determined that the preset fault ratio is greater than the fault ratio threshold, revising the invalid indicator value based on the indicator value of the data indicator reported before the fault; when it is determined that the preset fault ratio is less than or equal to the fault ratio threshold, revising the invalid data indicator value based on the indicator value of the data indicator reported in the previous report. Among them, the preset fault ratio threshold can be 50%.

[0078] For example, if the data indicator is the current of the actuator, and the preset fault ratio P corresponding to the current current is 60%, the time of the last fault can be located, and the historical status data reported before the last fault is determined according to the time of the last fault, and the current value in the historical status data reported before the last fault is used as the replacement value of the invalid current value in the current reported status data. If the preset fault ratio P corresponding to the current current is 30%, directly locate the previous reported historical status data, and use the current value in the previous reported historical status data as the replacement value of the invalid current value in the current reported status data.

[0079] If the data indicator does not have a preset fault rate, the invalid indicator value can be revised based on the fault impact level corresponding to the data indicator. Specifically, when the fault impact level corresponding to the data indicator is the highest value in history, the corresponding indicator value in the previously reported historical status data can be used as the replacement value for the invalid indicator value. When the fault impact level corresponding to the data indicator is the lowest value in history, the historical average value of the data indicator can be used as the replacement value for the invalid indicator value.

[0080] In another possible implementation, the replacement value of the invalid indicator value can also be determined according to the user's settings. It should be noted that, for each record of historical status data, the historical maximum value, historical minimum value, historical average value, most occurring value, value in the previous record, value in the next record, and least occurring value corresponding to each data indicator will be calculated. In this way, the user can select a data from the historical maximum value, historical minimum value, historical average value, most occurring value, value in the previous record, value in the next record, and least occurring value as the replacement value of the invalid indicator value based on his or her own experience.

[0081] In a possible implementation, the second preprocessing method may be to delete the invalid indicator value of the data indicator, or may be not to process the invalid indicator value, which is not specifically limited here.

[0082] S1300, preprocessing the index value of each data index according to the preprocessing method corresponding to each data index. The data preprocessing method has been described in step S1222, and will not be repeated here.

[0083] After the index values ​​of each data index in the currently reported status data are preprocessed, the preprocessing operation of the status data is completed. After the preprocessing operation is completed, it also includes the operations of data cleaning, data integration, data conversion, and data storage based on the big data processing technology.

[0084] The data cleaning process is the same as the preprocessing process and will not be described here. During the data integration process, the cleaned data is stored in the server through HDFS. During the data conversion process, the indicator value of each data indicator in the status data is converted into the fault impact degree. Then, when a fault occurs, the key indicators and non-key indicators are determined based on the fault impact degree corresponding to each data indicator, and the key indicators are recommended to the engineers who solve the fault to assist the engineers in judging and solving the fault. After the data conversion, the obtained status data and the fault impact degree corresponding to each indicator in the status data are stored as regular status data.

[0085] When preprocessing the status data of the robot, the present invention first determines the preprocessing method corresponding to each data indicator in the status data according to the fault data of the robot, and then preprocesses it according to the preprocessing method corresponding to each data indicator. In this way, the preprocessing method corresponding to each data indicator can be dynamically adjusted according to the fault change of the robot, so that the preprocessed status data meets the needs of dynamic fault diagnosis and analysis of the robot.

[0086] <Method Example>

[0087] Figure 2 FIG. 1 is a schematic flow chart showing an example of a method for preprocessing robot state data according to an embodiment of the present disclosure. Figure 2 As shown, the method comprises the following steps:

[0088] First, the robot reports the industrial robot status data through the network at a certain frequency. The status data reported at one clock frequency is called a record.

[0089] Second, traverse each data indicator in the status data. For the currently traversed data indicator x, obtain the corresponding fault frequency F and the fault indicator value R(x) of the data indicator x in each fault history data, use the above formula (2) to calculate the fault dynamic impact factor K(x) corresponding to the data indicator x, and calculate the dynamic fault ratio corresponding to the data indicator x according to the above formula (1). Obtain the historical indicator value H(x) (i.e., regular metadata) of the data indicator x in each historical status data, the preset fault ratio p corresponding to the data indicator x, and the data indicator threshold T(j) corresponding to each institution, and use the above formula (3) to calculate the fault impact degree f(x) corresponding to the data indicator x. After the traversal is completed, the fault impact degree f(x) corresponding to each data indicator in the status data is obtained.

[0090] Third, sort the data indicators according to the degree of impact of the fault, and take the data indicator of the set data with the greatest degree of impact as the key data indicator, and the remaining indicators as non-key indicators.

[0091] Fourth, traverse the indicator value of each data indicator in the state data. For the indicator value of the currently traversed data indicator x, determine whether it is an invalid indicator value. If it is an invalid indicator value and the data indicator x is a non-critical indicator, you can choose to delete or not process the invalid indicator; if it is an invalid indicator value and the data indicator x is a critical indicator, you can choose the above-mentioned automatic filling algorithm or user screening method to pre-process the invalid indicator value. The traversal is completed, and the pre-processing operation of the state data is completed.

[0092] Fifth, based on big data processing technology, the pre-processed state data is cleaned and integrated in turn, and the state data is converted into regular state data for storage.

[0093] When a fault occurs, the key indicators are determined based on the fault impact f(x) corresponding to each data indicator, and the key indicators are pushed to the engineers who are handling the fault to assist the engineers in fault analysis, fault solution formulation, and fault handling according to the fault solution. After the fault is handled, the fault frequency F, fault history data (i.e. fault history information), and preset fault proportion P corresponding to each data indicator are updated based on the current fault.

[0094] It can be seen from the present embodiment that each time a fault occurs, the fault frequency F, fault history data (i.e., fault history information) and preset fault proportion P corresponding to each indicator will be updated, thereby changing the degree of fault impact corresponding to each data indicator. As the degree of fault impact changes, the classification of data indicators will change, thereby achieving real-time adjustment of the preprocessing method corresponding to each data indicator according to the change of the robot fault, so that the processed status data can meet the needs of dynamic fault diagnosis and analysis of the robot.

[0095] <Device Example>

[0096] Figure 3 FIG. 1 is a schematic block diagram of a robot state data preprocessing device according to an embodiment of the present disclosure. Figure 3 As shown, the robot state data preprocessing device 100 includes:

[0097] The data acquisition module 110 is used to read the indicator value of each data indicator from the status data currently reported by the robot.

[0098] The processing method determination module 120 is used to determine the preprocessing method corresponding to each data indicator according to the fault data of the robot.

[0099] The preprocessing module 130 is used to preprocess the indicator value of each data indicator according to the preprocessing method corresponding to each data indicator.

[0100] In a possible implementation, the processing method determination module 120 includes a fault impact degree calculation submodule and a preprocessing method selection submodule. When determining the preprocessing method corresponding to each data indicator according to the robot's fault data:

[0101] The fault impact degree calculation submodule is used to calculate the fault impact degree corresponding to each data indicator based on the robot's fault data.

[0102] The preprocessing method selection submodule is used to determine the preprocessing method corresponding to each data indicator according to the fault impact degree corresponding to each data indicator.

[0103] In a possible implementation, the fault data includes at least one of a fault dynamic impact factor corresponding to each data indicator and a preset fault proportion, wherein the fault dynamic impact factor corresponding to each data indicator is calculated based on the fault frequency and fault history data corresponding to each data indicator.

[0104] In a possible implementation, the calculation formula for the fault impact degree is as follows:

[0105]

[0106] Where f(x) is the fault impact degree corresponding to the data indicator x, H(x i ) is the historical indicator value of data indicator x in the i-th historical status data, d is the corresponding fault dynamic impact factor of data indicator x, p is the preset fault proportion, and T(j) is the data indicator threshold corresponding to each mechanism of the robot.

[0107] In a possible implementation, the preprocessing method selection submodule includes an indicator classification subunit and a processing method selection subunit. When determining the preprocessing method corresponding to each data indicator according to the fault impact degree corresponding to the data indicator:

[0108] The indicator classification subunit is used to classify each data indicator into a critical data indicator and a non-critical data indicator according to the fault impact degree corresponding to each data indicator.

[0109] The processing method selection subunit is used to use the preset first preprocessing method as the preprocessing method corresponding to the key data indicator, and use the preset second preprocessing method as the preprocessing method for the non-key data indicator.

[0110] In a possible implementation, the first preprocessing method includes using a preset filling algorithm to fill the invalid indicator value of the data indicator. The preprocessing module 130 includes an invalid indicator determination submodule and an invalid data revision submodule. When the preset filling algorithm is used to fill the invalid indicator value of the data indicator:

[0111] The invalid indicator determination submodule is used to determine whether the indicator value of the data indicator is an invalid indicator value.

[0112] The invalid data revision submodule is used to revise the invalid indicator value based on a preset fault ratio corresponding to the data indicator when determining that the indicator value of the data indicator is an invalid indicator value.

[0113] In a possible implementation, the invalid data revision submodule includes a threshold determination subunit, a first revision subunit, and a second revision subunit. When revising the invalid indicator value based on a preset fault ratio corresponding to the data indicator:

[0114] The threshold determination subunit is used to determine whether a preset fault ratio is greater than a preset fault ratio threshold.

[0115] The first revision subunit is used to revise the invalid indicator value based on the indicator value of the data indicator reported before the failure when it is determined that the preset fault proportion is greater than the fault proportion threshold.

[0116] The second revision subunit is used to revise the invalid data indicator value based on the indicator value of the previously reported data indicator when it is determined that the preset fault ratio is less than or equal to the fault ratio threshold.

[0117] In one possible implementation, the second preprocessing method includes deleting invalid indicator values ​​of the data indicators.

[0118] <Equipment Embodiment>

[0119] Figure 4FIG. 1 is a schematic block diagram of a robot state data preprocessing device according to an embodiment of the present disclosure. Figure 4 As shown, the robot state data preprocessing device 200 includes a processor 210 and a memory 220 for storing executable instructions of the processor 210. The processor 210 is configured to implement any of the above robot state data preprocessing methods when executing the executable instructions.

[0120] Here, it should be noted that the number of processors 210 may be one or more. At the same time, in the robot state data preprocessing device 200 of the embodiment of the present disclosure, an input device 230 and an output device 240 may also be included. Among them, the processor 210, the memory 220, the input device 230 and the output device 240 may be connected through a bus or in other ways, which are not specifically limited here.

[0121] The memory 220 is a computer-readable storage medium that can be used to store software programs, computer executable programs, and various modules, such as the program or module corresponding to the robot state data preprocessing method of the embodiment of the present disclosure. The processor 210 executes various functional applications and data processing of the robot state data preprocessing device 200 by running the software program or module stored in the memory 220.

[0122] The input device 230 may be used to receive input numbers or signals. The signals may be key signals related to user settings and function control of the device / terminal / server. The output device 240 may include display devices such as display screens.

[0123] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or technical improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A method for preprocessing robot state data, characterized in that: include: Read the indicator value of each data indicator from the status data currently reported by the robot; Determine a preprocessing method corresponding to each of the data indicators according to the fault data of the robot; Preprocessing the indicator value of each of the data indicators according to the preprocessing method corresponding to each of the data indicators; When determining the preprocessing method corresponding to each of the data indicators according to the fault data of the robot, it includes: Calculating the degree of fault impact corresponding to each of the data indicators according to the fault data of the robot; According to the degree of fault impact corresponding to each of the data indicators, the preprocessing method corresponding to each of the data indicators is determined. Specifically, each of the data indicators is sorted according to the degree of fault impact, and the data indicator of the set data with the largest degree of fault impact is used as the key data indicator, and the remaining indicators are used as non-key indicators. The preset first preprocessing method is used as the preprocessing method corresponding to the key data indicator, and the preset second preprocessing method is used as the preprocessing method for the non-key data indicators.

2. The method according to claim 1, characterized in that The fault data includes at least one of a dynamic fault proportion and a preset fault proportion corresponding to each of the data indicators; The dynamic fault proportion corresponding to each of the data indicators is calculated based on the fault frequency and fault history data corresponding to each of the data indicators.

3. The method according to claim 2, characterized in that The calculation formula of the fault impact degree is as follows: Where f(x) is the fault impact degree corresponding to the data indicator x, H(x i ) is the historical indicator value of the data indicator x in the i-th historical status data, d is the dynamic fault ratio corresponding to the data indicator x, p is the preset fault ratio, and T(j) is the data indicator threshold corresponding to each mechanism of the robot.

4. The method according to claim 1, characterized in that: The first preprocessing method includes using a preset filling algorithm to fill the invalid indicator value of the data indicator; When the invalid indicator value of the data indicator is filled using a preset filling algorithm, it includes: Determining whether the indicator value of the data indicator is an invalid indicator value; When it is determined that the indicator value of the data indicator is an invalid indicator value, the invalid indicator value is revised based on a preset fault proportion corresponding to the data indicator.

5. The method according to claim 4, characterized in that When revising the invalid indicator value based on the preset fault proportion corresponding to the data indicator, it includes: Determine whether the preset fault ratio is greater than a preset fault ratio threshold; When it is determined that the preset fault proportion is greater than the fault proportion threshold, revising the invalid indicator value based on the indicator value of the data indicator reported before the fault; When it is determined that the preset fault proportion is less than or equal to the fault proportion threshold, the invalid data indicator value is revised based on the indicator value of the data indicator reported in the previous item.

6. The method according to claim 1, characterized in that The second preprocessing method includes deleting invalid indicator values ​​of the data indicators.

7. A robot state data preprocessing device, characterized in that: include: The data acquisition module is used to read the indicator value of each data indicator from the status data currently reported by the robot; A processing method determination module, used to determine a preprocessing method corresponding to each of the data indicators according to the fault data of the robot; A preprocessing module, used to preprocess the index value of each of the data indicators according to the preprocessing method corresponding to each of the data indicators; The processing method determination module, when determining the preprocessing method corresponding to each of the data indicators according to the fault data of the robot, is specifically used to: Calculating the degree of fault impact corresponding to each of the data indicators according to the fault data of the robot; According to the degree of fault impact corresponding to each of the data indicators, the preprocessing method corresponding to each of the data indicators is determined. Specifically, each of the data indicators is sorted according to the degree of fault impact, and the data indicator of the set data with the largest degree of fault impact is used as the key data indicator, and the remaining indicators are used as non-key indicators. The preset first preprocessing method is used as the preprocessing method corresponding to the key data indicator, and the preset second preprocessing method is used as the preprocessing method for the non-key data indicators.

8. A robot state data preprocessing device, characterized in that: include: processor; a memory for storing processor-executable instructions; Wherein, the processor is configured to implement the method described in any one of claims 1 to 6 when executing the executable instructions.

Images