Full-automatic plugging test machine control system and method

Abstract

The application relates to the technical field of intelligent control, in particular to a full-automatic plug-in test machine control system and method, which comprises a multi-dimensional data acquisition module, an abnormal state identification module, an action termination judgment module, a parameter correction module and a trajectory optimization execution module.In the application, through multi-sensor array fusion, the coordinated acquisition of multiple physical quantities during equipment operation is realized, vibration, temperature, path and plug-in force data are comprehensively calibrated by using a time axis, through frequency spectrum analysis and three-dimensional deviation comparison, subtle abnormalities and trend changes are dynamically revealed, running fluctuations and potential abnormalities are identified in real time, plug-in force reasonable intervals are automatically calculated, action duration and path direction are adjusted, and running parameters are continuously corrected, so that equipment abnormality early warning, accurate intervention and multi-parameter linkage regulation are realized in the whole process, the test continuity, execution sensitivity and prevention and control capability are improved, the misoperation is effectively inhibited and the wear is reduced, and the high-density intelligent test demand is met.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent control, in particular to a full-automatic plug-in test machine control system and method. BACKGROUND

[0002] The technical field of intelligent control involves the use of computers and automation technology to control and manage devices and systems. This field mainly covers the design and optimization of control systems, sensor technology, actuator control, data acquisition and processing, and the application of intelligent algorithms. Intelligent control technology can improve the automation level of devices, enhance the accuracy and reliability of production processes, and is widely used in industries, transportation, energy, medical care, and other fields. Key issues include modeling and analysis of control systems, design of control algorithms, real-time monitoring and feedback mechanisms of systems, and the goal is to improve the automation and efficiency of devices or systems through intelligent means.

[0003] Among them, the traditional full-automatic plug-in test machine control system refers to a device control system for testing plug-in performance. This system uses relay control, PLC control, analog circuit control and other means to operate and manage the device. The traditional system sets parameters manually and performs plug-in tests, but its automation level is low, and it cannot monitor and adjust parameters in real time, resulting in low operating efficiency, especially in high-frequency and large-scale test environments, which is prone to errors or tedious operations. In order to improve the test accuracy and operating efficiency, the traditional control system completes the automation of plug-in test through fixed program setting and hardware connection, but there are problems such as slow device response speed and complex maintenance.

[0004] The existing system only performs single-type parameter setting and action control based on hardware logic units such as relays or PLCs, lacks real-time collaborative monitoring and trend analysis means, and is prone to data lag, feedback blind area and untimely abnormality identification when facing high-speed or multi-batch testing. Static parameter operation is difficult to actively perceive actual plug-in force, displacement and temperature fluctuations, and the device cannot self-correct when encountering abnormalities or wear and deformation during operation, which is prone to misoperation, device damage and complex maintenance, and is difficult to adapt to intelligent adjustment requirements in frequent switching or high-strength environments. SUMMARY

[0005] In order to solve the technical problems that the existing system only performs single-type parameter setting and action control based on a hardware logic unit such as a relay or a PLC, lacks multivariate real-time collaborative monitoring and trend analysis means, is prone to data lag, feedback blind area and untimely abnormality identification when facing high-speed or multi-batch testing, is difficult to actively perceive actual plug-in force, displacement and temperature fluctuation by using static parameter operation, cannot self-correct when encountering an abnormality or wear deformation during equipment operation, is prone to misoperation, equipment damage and complex maintenance, and is difficult to adapt to intelligent adjustment requirements in a frequently switched or high-strength environment, embodiments of the present application provide a full-automatic plug-in test machine control system and method. The technical solution is as follows:

[0006] In one aspect, a full-automatic plug-in test machine control system is provided, which comprises:

[0007] A multi-dimensional data acquisition module acquires time sequence data of acceleration, displacement change, temperature gradient, three-dimensional path coordinates, plug-in force and contact duration through a sensor array, aligns vibration, temperature, trajectory and force data according to time stamps, generates a synchronous data set, and delivers the synchronous data set to an abnormal state identification module;

[0008] The abnormal state identification module receives the synchronous data set, uses fast Fourier transform to process a vibration amplitude sequence to extract frequency characteristics, calculates a period amplitude difference value and a temperature change amount, performs comparison, combines trajectory offset amount comparison, generates a three-dimensional deviation set, and delivers the three-dimensional deviation set to an action termination judgment module;

[0009] The action termination judgment module receives the three-dimensional deviation set, scans and identifies vibration super-stable bands, temperature super-thermal stable ranges and trajectory offset super-continuous range time slices, sends a stop instruction when three continuous time slices meet the conditions, generates an abnormal trigger parameter, and delivers the abnormal trigger parameter to a parameter correction module;

[0010] The parameter correction module receives the abnormal trigger parameter, adjusts a plug-in force safety interval based on a vibration amplitude average value, corrects a stay duration according to a temperature change slope, calculates a path fine-tuning vector according to a trajectory offset direction, generates a corrected parameter set, and delivers the corrected parameter set to a trajectory optimization execution module.

[0011] As a further scheme of the present application, the synchronous data set comprises a data time alignment result, a signal synchronization accuracy index and a time sequence integrity check, the three-dimensional deviation set comprises a frequency spectrum characteristic difference value, a temperature gradient change amplitude and a spatial trajectory deviation amount, the abnormal trigger parameter comprises a continuous abnormal state judgment identifier, a device protection start signal and an abnormal action control code, and the corrected parameter set comprises a plug-in force safety threshold interval, a temperature change rate correction coefficient and a trajectory offset vector adjustment amount.

[0012] As a further scheme of the present application, the multi-dimensional data acquisition module comprises:

[0013] The vibration temperature acquisition sub-module acquires time sequence data of acceleration, displacement change, temperature gradient, three-dimensional path coordinates, plugging force and contact duration through a sensor array, aligns the acceleration and temperature gradient data based on a time stamp, calculates the correlation of acceleration and temperature change, and generates a warming correlation coefficient;

[0014] The trajectory data generation sub-module, based on the warming correlation coefficient, calls displacement change data, calculates displacement increments and direction vectors for each time period, applies coordinate transformation, normalizes displacement change amounts in time sequence, and generates three-dimensional path change amounts;

[0015] The mechanical characteristic alignment sub-module, according to the three-dimensional path change amounts, calls plugging force and contact duration data, calculates the increments of plugging force change and the relative changes of contact duration, aligns the data based on the data relationship between plugging force and contact duration, and generates a synchronous data set.

[0016] As a further scheme of the present application, the warming correlation coefficient refers to a numerical parameter calculated by a correlation analysis algorithm based on acceleration data and temperature gradient data of the same time stamp;

[0017] The normalization processing refers to using a range normalization method to linearly transform original three-dimensional path change amounts to a unified interval;

[0018] The three-dimensional path change amounts refer to vectors composed of displacement increments in X, Y and Z three-axis directions respectively calculated according to displacement change data at time intervals.

[0019] As a further scheme of the present application, the abnormal state recognition module comprises:

[0020] The frequency characteristic extraction sub-module, based on the synchronous data set, extracts each sampling point value of the vibration amplitude sequence and establishes an amplitude array in time sequence, calculates a frequency corresponding amplitude value sequence using fast Fourier transform, summarizes the amplitude value proportion according to frequency distribution, and generates a frequency amplitude value distribution coefficient;

[0021] The frequency amplitude value distribution coefficient refers to a parameter obtained by counting the proportion of the amplitude value in a frequency range in the total amplitude value after the vibration signal is decomposed by FFT;

[0022] The amplitude-temperature difference calculation sub-module calculates and correspondingly arranges the continuous period amplitude difference and temperature difference according to the frequency amplitude value distribution coefficient and the temperature sequence at the same time in the synchronous data set, and integrates them in time sequence to obtain a period amplitude difference-temperature difference ratio;

[0023] The period amplitude difference-temperature difference ratio refers to the ratio of the change amount of the vibration amplitude in the continuous period to the change amount of the temperature in the same period;

[0024] The offset ratio generation sub-module calls the period amplitude difference temperature difference ratio and the track offset sequence, constructs an offset trend and a ratio curve for the same time period, extracts a corresponding amplitude sequence, performs amplitude normalization matching according to a time sequence, maps to a three-dimensional coordinate, collects coordinate points, and generates a three-dimensional deviation set.

[0025] The amplitude normalization matching refers to a processing method of performing corresponding matching according to the same time sequence after the amplitude sequence is mapped to a unified interval by using the maximum and minimum normalization method.

[0026] As a further scheme of the application, the action termination judgment module comprises:

[0027] The deviation recognition sub-module obtains the acceleration deviation, the temperature deviation and the track offset value in the three-dimensional deviation set, judges whether each time slice exceeds the corresponding threshold according to the vibration recognition threshold, the temperature recognition threshold and the track offset threshold, extracts abnormal deviation time slice data, and generates an abnormal deviation amplitude interval.

[0028] The vibration recognition threshold is an upper limit of the acceleration absolute value or the change amount, and the time slice exceeding the value is determined as a vibration anomaly.

[0029] The temperature recognition threshold is a critical value set for the temperature deviation, and the time slice exceeding the temperature difference is determined as a temperature anomaly.

[0030] The track offset threshold is a maximum allowed deviation distance of the track point from the expected track, and the time slice exceeding the value is determined as a track anomaly.

[0031] The stable region scanning sub-module calls the vibration amplitude value, the temperature change range and the track offset rate in the corresponding time slice based on the abnormal deviation amplitude interval, filters the time slices meeting the stable condition according to the vibration stable judgment interval, the heat stable interval and the track continuous range, counts the number of continuously appearing groups, and generates the number of continuous stable segments.

[0032] The instruction generation sub-module judges whether there is a time period in which three adjacent time slices meet the three stable conditions at the same time according to the number of continuous stable segments, and if so, extracts the corresponding deviation data and generates an abnormal trigger parameter.

[0033] As a further scheme of the application, the parameter correction module comprises:

[0034] The abnormal recognition sub-module obtains the vibration amplitude data, the temperature acquisition data and the track offset vector value in the abnormal trigger parameter, calculates the average amplitude of the vibration time sequence, obtains the temperature change value per unit time to calculate the slope, calculates the offset direction according to the track point displacement, and generates an abnormal feature judgment value.

[0035] The abnormal feature judgment value refers to a judgment result for quantifying the abnormal degree of the current state after comprehensive analysis of parameters such as vibration amplitude average value, temperature change slope and trajectory deviation direction;

[0036] The safety interval adjustment submodule calls the vibration amplitude average value in the abnormal feature judgment value, sets the upper and lower limits of the initial safety interval of the plug-in force, calculates the percentage correction amplitude, adjusts the boundary value, and generates the plug-in force correction interval;

[0037] The correction set generation submodule calls the boundary value of the plug-in force correction interval, combines the temperature change slope and the trajectory deviation direction scalar in the abnormal feature judgment value, adjusts the residence time according to the temperature slope, sets the path adjustment vector according to the deviation direction, and generates a correction parameter set.

[0038] As a further scheme of the application, the trajectory optimization execution module receives the correction parameter set, smoothes the trajectory data using a spline interpolation function, re-plans the plug-in path sequence, adjusts the speed curve and contact time sequence, and executes the new trajectory three-dimensional deviation state of the corrector to generate an optimized trajectory execution instruction;

[0039] The optimized trajectory execution instruction includes spline interpolation curve parameters, speed curve shape adjustment parameters and contact time sequence optimization arrangement.

[0040] As a further scheme of the application, the trajectory optimization execution module includes:

[0041] The trajectory reconstruction submodule uses a spline interpolation function to interpolate the three-dimensional coordinates of the plug-in nodes in the original trajectory data based on the position offset and angle change value in the correction parameter set, recalculates the time interval and position change between nodes, adjusts the smoothness of the trajectory curve, and obtains a trajectory smoothness and continuity value;

[0042] The trajectory smoothness and continuity value is a quantitative index for measuring whether the speed and acceleration change between trajectory nodes are continuous and smooth;

[0043] The path adjustment submodule rearranges the path order between trajectory nodes according to the trajectory smoothness and continuity value and the speed limit threshold in the correction parameter set, calls the spatial distance between trajectory nodes and the contact order, adjusts the speed rhythm of the path based on the speed change, and generates a path order matching deviation rate;

[0044] The path order matching deviation rate represents the deviation proportion of the node spatial order after trajectory reconstruction relative to the original path;

[0045] The speed limit threshold is the maximum safe motion speed allowed in the trajectory planning process;

[0046] The error correction submodule matches the deviation rate with the interpolated trajectory data according to the path sequence, performs difference comparison on the three-dimensional trajectory deviation degree according to the current position deviation and the speed trajectory deviation data of the actuator, screens the trajectory nodes exceeding the speed change reference value, and generates an optimized trajectory execution instruction.

[0047] The speed change reference value is a reference threshold for determining whether the speed change of adjacent trajectory nodes is abnormal.

[0048] On the other hand, the full-automatic plug-in test machine control method is executed based on the full-automatic plug-in test machine control system, and includes the following steps:

[0049] S1: Collecting acceleration, displacement change, temperature gradient, three-dimensional path coordinates, plug-in force and contact duration time sequence data through a sensor array, aligning vibration, temperature, trajectory and force data according to time stamp, and generating a synchronous data set;

[0050] S2: Calling the synchronous data set, using fast Fourier transform to process vibration amplitude sequence to extract frequency characteristics, calculating and comparing period amplitude difference and temperature change, combining trajectory deviation comparison to generate a three-dimensional deviation set;

[0051] S3: Receiving the three-dimensional deviation set, scanning and identifying vibration super-stable band, temperature super-thermal stable range, trajectory deviation super-continuous range time slice, sending a stop instruction if the continuous three time slices meet the conditions, and generating an abnormal trigger parameter;

[0052] S4: Receiving the abnormal trigger parameter, adjusting the plug-in force safety interval based on the average value of the vibration amplitude, correcting the stay time according to the temperature change slope, calculating the path fine-tuning vector according to the trajectory deviation direction, and generating a correction parameter set;

[0053] S5: Calling the correction parameter set, using a spline interpolation function to perform smoothing operation on the three-dimensional path, re-planning the plug-in path sequence, calling the speed curve to adjust the speed parameter of the operation, rearranging the contact time sequence to obtain the corresponding plug-in path, verifying the three-dimensional deviation state of the corrected trajectory, and outputting an optimized trajectory execution instruction.

[0054] The technical scheme provided by the embodiment of the application has at least the following beneficial effects:

[0055] Based on multi-sensor array fusion, the device running in the multiple physical quantity cooperative collection is realized, and the data such as vibration, temperature, path and plugging force and the like which are related to each other are comprehensively calibrated with time axis, through frequency spectrum analysis and three-dimensional deviation comparison, subtle abnormalities and trend changes are dynamically revealed, running fluctuations and potential abnormal states are identified in real time, deviation parameters are automatically calculated to obtain reasonable plugging force interval, action duration and path direction are adjusted, running parameters are continuously corrected, and the device abnormality early warning, precise intervention and multi-parameter linkage adjustment are realized in the whole process, the test continuity, execution sensitivity and abnormal prevention and control ability are significantly improved, the misoperation is effectively inhibited and the equipment wear is reduced, and the intelligent matching demand under the high-density test environment is met. BRIEF DESCRIPTION OF DRAWINGS

[0056] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0057] Figure 1 is a schematic diagram of the full-automatic plugging test machine control system provided by the embodiments of the present application;

[0058] Figure 2 is a system framework schematic diagram of the present application;

[0059] Figure 3 is a flow chart of the multi-dimensional data acquisition module in the present application;

[0060] Figure 4 is a flow chart of the abnormal state recognition module in the present application;

[0061] Figure 5 is a flow chart of the action termination judgment module in the present application;

[0062] Figure 6 is a flow chart of the parameter correction module in the present application;

[0063] Figure 7 is a flow chart of the trajectory optimization execution module in the present application;

[0064] Figure 8 is a flow chart of the full-automatic plugging test machine control method provided by the embodiments of the present application. DETAILED DESCRIPTION

[0065] The technical solutions in the present application will be described below in combination with the drawings.

[0066] In the embodiments of the present application, the words such as "example", "for example" are used to represent an example, illustration, or description. Any embodiment or design scheme described as "example" in the present application should not be interpreted as more preferred or more advantageous than other embodiments or design schemes. Rather, the word "example" is intended to present the concept in a specific manner. In addition, in the embodiments of the present application, the meaning expressed by "and / or" can be both, or can be one of the two.

[0067] In the embodiments of the present application, "image" and "picture" can be used interchangeably at times, and it should be pointed out that the meanings expressed are consistent when the distinction is not emphasized. "Of", "corresponding" and "corresponding" can be used interchangeably at times, and it should be pointed out that the meanings expressed are consistent when the distinction is not emphasized.

[0068] In the embodiments of the present application, sometimes the subscript such as W1 can be written in the form of non-subscript such as W1, and the meanings expressed are consistent when the distinction is not emphasized.

[0069] In order to make the technical problems, technical schemes and advantages to be solved by the present application more clear, the following will be described in detail in combination with the drawings and specific embodiments.

[0070] The embodiments of the present application provide a full-automatic plug-in test machine control system, as shown in the full-automatic plug-in test machine control system schematic diagram, the system comprises: Figures 1-2

[0071] A multi-dimensional data acquisition module acquires time sequence data of acceleration, displacement change, temperature gradient, three-dimensional path coordinates, plug-in force and contact duration through a sensor array, aligns vibration, temperature, trajectory and force data according to time stamp, generates a synchronous data set, and delivers the synchronous data set to an abnormal state identification module;

[0072] The abnormal state identification module receives the synchronous data set, adopts fast Fourier transform to process a vibration amplitude sequence to extract frequency characteristics, calculates a period amplitude difference value and a temperature change amount and performs comparison, combines with trajectory offset amount comparison, generates a three-dimensional deviation set, and delivers the three-dimensional deviation set to an action termination judgment module;

[0073] The action termination judgment module receives the three-dimensional deviation set, scans and identifies vibration super-stable bands, temperature super-thermal stable ranges, and trajectory offset super-continuous range time slices, sends a stop instruction when the three time slices continuously meet the conditions, generates an abnormal trigger parameter, and delivers the abnormal trigger parameter to a parameter correction module;

[0074] ​The parameter correction module receives the abnormal trigger parameter, adjusts the plug-in force safety interval based on the vibration amplitude average value, corrects the stay time according to the temperature change slope, calculates the path fine adjustment vector according to the trajectory deviation direction, generates a corrected parameter set, and transmits the corrected parameter set to the trajectory optimization execution module;

[0075] The trajectory optimization execution module receives the corrected parameter set, smoothes the trajectory data by using a spline interpolation function, re-plans the plug-in path sequence, adjusts the speed curve and the contact time sequence, and executes the new trajectory three-dimensional deviation state by the corrector, and generates an optimized trajectory execution instruction.

[0076] The synchronization data set includes data time alignment results, signal synchronization accuracy indicators, and time sequence integrity verification, the three-dimensional deviation set includes spectrum feature difference values, temperature gradient change amplitudes, and spatial trajectory deviation amounts, the abnormal trigger parameter includes continuous abnormal state determination identifiers, device protection start signals, and abnormal action control codes, the corrected parameter set includes plug-in force safety threshold intervals, temperature change rate correction coefficients, and trajectory deviation vector adjustment amounts, and the optimized trajectory execution instruction includes spline interpolation curve parameters, speed curve shape adjustment parameters, and contact time sequence optimization arrangements.

[0077] Specifically, as shown in Figure 2 , 3 The multi-dimensional data acquisition module includes:

[0078] The vibration temperature acquisition sub-module acquires time sequence data of acceleration, displacement change, temperature gradient, three-dimensional path coordinates, plug-in force, and contact duration through a sensor array, aligns acceleration and temperature gradient data based on time stamps, calculates the correlation between acceleration and temperature change, and generates a temperature correlation coefficient.

[0079] The vibration and temperature acquisition submodule collects time-series data on acceleration, displacement change, temperature gradient, three-dimensional path coordinates, insertion / extraction force, and contact duration through a sensor array. First, it calls a six-axis inertial measurement unit (IMU) to record real-time three-dimensional acceleration and angular velocity, for example, using an ADXL355 series accelerometer at a sampling frequency of 100Hz for 60 seconds, resulting in 6000 sets of acceleration data. Then, it uses a linear displacement sensor (such as an LVDT) to record displacement change data at a sampling frequency of 50 times per second, matching the 20mm displacement sampling during insertion / extraction, for a total of 1000 sets of displacement data. Temperature gradient information is collected by multiple thermocouples (such as K-type thermocouples) distributed around the insertion / extraction contact, positioned at the pin center, socket edge, and connector housing, with a sampling frequency of 1Hz, recording the temperature differences at different points throughout the process. Simultaneously, a three-dimensional path is acquired through a three-dimensional trajectory sensor (such as an optical tracking module). The coordinates are recorded to show the changes in path coordinates during displacement. The starting coordinates are set to (0, 0, 0), and the displacement increment vector is accumulated point by point to generate the path coordinates. The insertion and extraction force is recorded at the contact point using an embedded force sensor (such as a piezoelectric sensor), with a maximum measurement range of 0–50N and a sampling frequency of 200Hz. The contact duration is detected by a system timer and trigger, recording the start and end times and converting them to seconds. For example, if the initial contact time is 2.34s and the end time is 2.91s, the contact time is 0.57s. Then, based on the timestamp information of the recorded data, the collected data is uniformly organized by second-level timestamps. A bidirectional matching search mechanism is used to pair acceleration and temperature gradient data according to the closest timestamp, such as pairing acceleration data at timestamp 2.500s with temperature data at timestamp 2.498s. The total number of matching pairs is determined with an error of less than 0.005s. Then, each pair of data is traversed, and each pair of acceleration data is retrieved. With temperature gradient The values ​​are normalized and their Pearson correlation coefficients are calculated using the formula:

[0080] ;

[0081] in and Here, n represents the mean values ​​of acceleration and temperature gradient, respectively, and n is the number of paired data sets. A practical example is provided below: Assume the acceleration in the five sampled data sets is [0.98, 1.02, 1.00, 1.01, 0.99]. The temperature gradient is [2.5, 2.6, 2.4, 2.5, 2.3]. After calculating the mean, they are respectively , Substitute the terms into the formula and perform the calculations one by one:

[0082] ;

[0083] The correlation coefficient result is 0.444, which is between the interval [0.3, 0.6], and according to the definition of the actual standard, the value is classified as "moderate correlation", so this data set represents the existence of a moderate linear correlation between acceleration and temperature gradient in the test stage, and the result is the warming correlation coefficient; The threshold setting interval of the correlation coefficient here is: weak correlation [0-0.3), moderate correlation [0.3-0.6), strong correlation [0.6-1.0], and the reference source is set to the average correlation degree statistical distribution range of the original multi-batch test data, and combined with the IEEE engineering experiment standard for classification and division.

[0084] Table 1: Acceleration and temperature gradient collection example table

[0085]

[0086] As shown in Table 1, the acceleration and temperature gradient data corresponding to 5 groups of key time points in a certain plug-in and plug-out process are listed, and the moderate linear correlation warming correlation coefficient value is obtained after normalization and correlation operation.

[0087] The trajectory data generation submodule, based on the warming correlation coefficient, calls the displacement change data, calculates the displacement increment and direction vector of each time period, applies coordinate transformation, and normalizes the displacement change according to the time sequence to generate three-dimensional path change;

[0088] The trajectory data generation submodule based on the above obtained warming correlation coefficient first calls the displacement change data, reads the displacement change corresponding to each timestamp in the time sequence database recorded by the system, and the data structure is defined as a three-dimensional vector form , each vector represents the three-dimensional position increment between adjacent two timestamps, for example, between time points and , the displacement coordinate change recorded by the detector changes from to , then the three-dimensional increment in this time period is , then the increment vector is unit vectorized in three-dimensional space, that is, it is divided by the vector length to get the direction vector, and the calculation process of the length is:

[0089] ;

[0090] The direction vector is , which represents the normalized direction of the motion direction in this time period, and then the normalization process is called to uniformly convert the three-dimensional displacement change to the interval [0, 1], and the range normalization method is used, that is, the displacement change of each axis is transformed by the following linear transformation formula:

[0091] ;

[0092] Suppose that the minimum value of the X-direction displacement is 9.5 mm, the maximum value is 19.5 mm, and the current displacement value is 13.1 mm in the entire test period, then the normalized value is:

[0093] ;

[0094] Similarly, the normalized coordinate triplet is obtained after processing the coordinate changes in the Y and Z directions: Through this process, the original path change value in the time period is unified to the standard scale interval, thereby eliminating the influence of the absolute dimensional difference of the path coordinates; further, in order to make the normalized result have dynamic adjustment capability, the heating correlation coefficient obtained in the previous section is introduced as an adjustment factor to adjust the weight of each displacement change value. The operation mode is to take the value of the heating correlation coefficient as a weighting factor and multiply it into each normalized displacement result, for example, the normalized result of the current path point is , and the corresponding , then the weighted coordinate is:

[0095] ;

[0096] The weighted normalized vector of each time period is sequentially stored in the path array and accumulated, that is, the overall three-dimensional path change is generated in an accumulative manner, for example, the starting point is the origin , the first time period is , and the second period is , then the second path coordinate is , and so on to obtain the complete path sequence; the path array represents the standardized motion trajectory of the connector terminal in the three-dimensional space during the entire insertion and extraction action, and the array serves as an important coordinate reference for subsequent alignment with the mechanical characteristic data.

[0097] The mechanical characteristic alignment submodule calls the insertion and extraction force and contact duration data according to the three-dimensional path change, calculates the increment of the insertion and extraction force change and the relative change of the contact duration, performs data alignment based on the data relationship between the insertion and extraction force and the contact duration, and generates a synchronous data set;

[0098] The mechanical characteristic alignment submodule first calls the insertion and extraction force and contact duration data according to the three-dimensional path change calculated in the previous section, arranges the insertion and extraction force change and the contact duration at each time point in chronological order to ensure that the data are aligned with the path data. For example, at the time stamp , the system records the insertion and extraction force as , and the contact duration as ​​, the three-dimensional path change amount at the same time point is Next, for the alignment data, the change increment of the plug-in force and the relative change of the contact duration are calculated, and the plug-in force change increment The plug-in force difference between the adjacent two time points can be obtained:

[0099] ;

[0100] Similarly, the change increment of the contact duration The duration difference between adjacent time points is calculated:

[0101] ;

[0102] Next, based on the relationship between the plug-in force and the contact duration, the data is aligned, and the time alignment accuracy between the data is ensured to be 0.01 seconds through dynamic time warping (DTW) or interpolation method. At this time, the plug-in force and the contact time are synchronized with the path change amount at the same time interval, ensuring that the physical characteristics and trajectory changes in the entire plug-in process are reflected at the same time, and forming a final synchronized data set in time sequence for subsequent analysis and prediction. For example, at the time stamp , the synchronized data set formed contains the following information:

[0103] ;

[0104] In this way, the synchronization and correlation between the data can be ensured, thereby providing reliable input data for subsequent analysis, modeling and prediction.

[0105] Specifically, as shown in Figure 2 , 4 , the abnormal state recognition module includes:

[0106] The frequency feature extraction submodule extracts the value of each sampling point of the vibration amplitude sequence based on the synchronized data set and establishes an amplitude array in time sequence, calculates the frequency corresponding amplitude value sequence using fast Fourier transform, summarizes the amplitude value proportion according to frequency distribution, and generates a frequency amplitude distribution coefficient.

[0107] The frequency feature extraction submodule based on the synchronized data set first extracts the vibration amplitude sequence associated with the plug-in action process, arranges the vibration amplitude value of each sampling point in time stamp order to form a one-dimensional array, assuming that the sampling period is 0.01 seconds, and 500 sampling points of data collection are completed within 5 seconds, then the amplitude array is constructed in the form of , where each item represents the time point the instantaneous vibration amplitude of the sequence obtained by quantifying the maximum acceleration amplitude detected by the vibration sensor at each time point, for example, the value detected at time point , and so on, to form a continuous vibration response data during the entire plug-in and plug-out process. Then, a fast Fourier transform operation is performed on the amplitude array to convert the original time domain signal to the frequency domain to obtain the response amplitude at the frequency point. The transformation process needs to calculate the complex modulus of the real part and the imaginary part and record the corresponding frequency interval. In this embodiment, the frequency distribution is from 0 Hz to 250 Hz, and the corresponding amplitude sequence is obtained after conversion , where each item represents the signal energy at the corresponding frequency

[0108] . Further, the percentage of the total amplitude energy occupied by the frequency band is calculated, and the frequency amplitude distribution coefficient is extracted through the amplitude ratio of the frequency distribution. This process needs to divide the frequency band into several subintervals, and calculate the total energy ratio of each interval, for example, dividing 0-250 Hz into five subintervals:

[0109] The ratio of the total amplitude sum in each subinterval to the total amplitude sum is calculated, for example, the total amplitude sum of the first interval is , the total frequency amplitude sum is , and the frequency amplitude distribution coefficient of the interval is . Similarly, the percentage of all frequency intervals is calculated, and finally represented in the form of a vector . The frequency amplitude distribution coefficient vector is used as the frequency domain feature input. The following table lists the frequency, corresponding amplitude, and calculated percentage of the frequency band under the example data, which is used to show the typical results in the execution process of this step.

[0110] Table 2: Frequency amplitude distribution table

[0111]

[0112] As shown in Table 2, the frequency bands with large amplitude sums are concentrated in 100-150 Hz and 200-250 Hz, indicating that the main energy in the plug-in and plug-out action is concentrated in these two frequency bands. The amplitude coefficient of the 150-200 Hz band is 0.201, which belongs to the normal energy band, while the coefficient of the 0-50 Hz band is 0.153, which is slightly lower than the subsequent frequency bands. If the coefficient of this band is lower than 0.05 in the remaining samples, it should be judged as a low-frequency response area, and further classified as an edge feature frequency band.

[0113] ​​​​The amplitude-temperature difference calculation sub-module calculates the amplitude difference and the temperature difference in a continuous period according to the frequency amplitude distribution coefficient and the temperature sequence at the same time in the synchronized data set, arranges the amplitude difference and the temperature difference correspondingly, integrates in time sequence, and obtains the amplitude difference-temperature difference ratio in a period;

[0114] After the frequency amplitude distribution coefficient is extracted by the amplitude-temperature difference calculation sub-module, the synchronized time sequence temperature data is called, the frequency amplitude vector at each time point is matched with the temperature value collected at the same time point, for example, at the time point , the frequency amplitude distribution coefficient vector is , the corresponding temperature value is , then the system calls the corresponding parameters of the next sampling period, for example, the frequency amplitude vector at the time point is , the corresponding temperature is , the amplitude and the temperature at the two time points are compared, and the change of the frequency amplitude sum is calculated:

[0115]

[0116] The corresponding temperature difference is , and the amplitude difference-temperature difference ratio in the period is obtained:

[0117] ;

[0118] Similarly, in the whole sampling period, the operation is performed on the adjacent time periods, the amplitude difference-temperature difference ratio sequence between the adjacent sampling points is extracted, a set of array data sorted by time is formed, and the data structure is:

[0119] ;

[0120] Each value can specifically identify the sensitivity of the frequency response intensity at different stages to the temperature change, and it should be noted that in the calculation process, the period in which the temperature change is less than the set threshold value is excluded from the ratio calculation, so as to avoid the denominator tending to 0, resulting in that the calculation result is not comparable, and the threshold value is set to in the embodiment, when the temperature difference is lower than the value, the system sets the ratio to be invalid or directly skips, for example, when the temperature change of a period is , the system does not perform the ratio operation of the period, and when the complete amplitude difference-temperature difference ratio sequence is constructed, the valid ratio points are reserved and stored with the original time stamp, forming a ratio curve sequence with time attribute, and the final result is an amplitude-temperature ratio array changing with time, which is used for subsequent mapping and fusion analysis with the track offset trend.

[0121] The offset ratio generation submodule calls the cycle amplitude difference temperature difference ratio and the trajectory offset sequence, constructs the offset trend and the ratio curve for the same time period, extracts the corresponding amplitude sequence, performs amplitude normalization matching according to the time sequence, and maps to a three-dimensional coordinate, collects the coordinate points, and generates a three-dimensional deviation set;

[0122] The amplitude normalization matching refers to a processing method of corresponding matching after mapping the amplitude sequence to a unified interval by using the maximum and minimum normalization method;

[0123] The offset ratio generation submodule accesses the amplitude difference temperature difference ratio sequence obtained by calculation and the trajectory offset sequence obtained in the early stage, performs point-by-point matching operation on the two data sets in the same time period, and first aligns the ratio corresponding to the time point in the amplitude temperature ratio sequence with the trajectory offset , for example, at time point , the amplitude temperature ratio is , and the trajectory offset is , at subsequent time point , the corresponding ratio is , and the offset is , the system combines the ratio and the offset data to form a binary sequence , and then performs maximum and minimum normalization processing on the amplitude sequences of the ratio and the offset, specifically, the ratio is mapped to the interval [0, 1], and the normalization method is to subtract the minimum value of the sequence from the current value and divide by the difference between the maximum value and the minimum value, for example, if the maximum value of the ratio sequence is 0.078 and the minimum value is 0.032, then the normalized ratio 0.054 is:

[0124] ;

[0125] The offset is normalized in the same way to obtain a normalized ratio sequence and a normalized offset sequence, and the system aligns and matches the two sets of normalized data to form a normalized mapping pair, for example, the normalized ratio is 0.478, and the normalized offset is 0.513, then the normalized matching result at this time point is a three-dimensional coordinate point , wherein the time is input as the horizontal axis, and the complete time-amplitude-offset three-dimensional mapping set is sequentially constructed, if the normalized ratio and the normalized offset change trend direction are inconsistent in a certain time period, for example, the ratio increases and the offset decreases, the system needs to record this section as a reverse offset interval, further analyzes the relationship between the occurrence time and the mechanical behavior, and finally the time point matching result is collected to form a complete three-dimensional offset set, and the data structure is , wherein is the normalized ratio, is the normalized offset, which is used for subsequent data support.

[0126] Specifically, as shown in Figure 2 , 5 , the action termination judgment module includes:

[0127] The deviation identification submodule obtains the acceleration deviation, temperature deviation, and trajectory offset value in the three-dimensional deviation set, judges whether each time slice exceeds the corresponding threshold according to the vibration identification threshold, temperature identification threshold, and trajectory offset threshold, extracts abnormal deviation time slice data, and generates an abnormal deviation amplitude interval.

[0128] The deviation identification submodule extracts the acceleration deviation value, temperature deviation value, and trajectory offset value corresponding to each time slice from the three-dimensional deviation set, performs an independent threshold judgment operation on each parameter. First, the system calls the acceleration deviation sequence , the temperature deviation sequence , and the trajectory offset sequence after timestamp alignment, sets the vibration identification threshold to , the temperature identification threshold to , and the trajectory offset threshold to , and then performs the following judgment on each time slice : if , it is marked that the time slice has an abnormal vibration deviation, if , it is marked that the time slice has an abnormal temperature deviation, and if , it is marked that the time slice has a trajectory offset anomaly. In specific implementation, for example, for time slice , if the three data are , , , the system first compares , which satisfies the vibration deviation anomaly condition, then compares , which satisfies the temperature deviation anomaly condition, but since , it does not satisfy the trajectory offset anomaly condition, the time slice is marked as an abnormal type of “vibration + temperature”. The system iterates through the time slices and executes the above three types of judgment logic for each segment to obtain the time index set of each type of anomaly, further generates a three-type anomaly marking matrix, and then calculates the anomaly amplitude interval according to the anomaly type and amplitude value marked for each time slice. For example, the amplitude is defined as the maximum relative deviation in the three types of anomalies. If the corresponding time slice , , , the corresponding relative deviation is:

[0129] , , ;

[0130] The maximum of the three is 0.5, so the abnormal amplitude mark is "0.5", and the corresponding segment enters the abnormal amplitude interval "0.5 level". By performing such calculations on the segments, the system constructs a complete set of abnormal amplitude interval , wherein represents an abnormal type combination, such as "vibration + temperature + trajectory", represents the corresponding maximum relative deviation amplitude value, such as "0.5". Finally, the time slices that satisfy the threshold exceeding condition of any one type of anomaly are extracted to form an abnormal deviation time slice sequence, which is used for stability analysis and trigger logic recognition in the next stage. The data used in this step is shown in the following example:

[0131] Table 3: Abnormal deviation judgment data table

[0132]

[0133] As shown in Table 3, by comparing the difference ratio of each type of parameter and its corresponding identification threshold, the maximum relative deviation value is obtained and used to archive the abnormal amplitude level of the time slice, forming an abnormal deviation amplitude interval set, which provides data basis for subsequent stability identification. The results show that the system calculates the deviation ratio of acceleration, temperature and trajectory value one by one, and obtains the abnormal level through the maximum value, realizing the abnormal identification and partition of time slice.

[0134] The stable area scanning submodule, based on the abnormal deviation amplitude interval, calls the vibration amplitude value, temperature change range and trajectory offset rate in the corresponding time slice, and filters the time slices that meet the stability conditions according to the vibration stability judgment interval, heat stability interval and trajectory continuous range, counts the number of continuous occurrence groups, and generates the number of continuous stable segments;

[0135] After receiving the abnormal deviation amplitude interval, the stable area scanning submodule calls the vibration amplitude value, temperature change range and trajectory offset rate in the corresponding time slice for each time slice marked as abnormal, and respectively corresponds to the acceleration absolute value , temperature gradient , displacement first-order difference value , and compares them with the three stability judgment criteria in turn. The vibration stability judgment interval is set to , the temperature stability interval is , and the trajectory continuous range is set to In specific implementation, the system first obtains the acceleration values of the current time point and its adjacent time points for each time slice, and calculates the vibration amplitude by using a sliding window method, i.e. taking the maximum value of the absolute values of the three points in the window as the vibration index of the current segment. If the maximum value is less than 0.35, it meets the vibration stability condition. For example, the time slice , the acceleration of the previous and next points is The maximum value of the three values ​​is 0.31, which is less than the upper limit for judgment, thus satisfying the vibration stability condition; the temperature change range is calculated by taking the absolute value of the difference between the temperature values ​​at adjacent time points. For example, the temperatures before and after the time slice are respectively... and Corresponding The difference in value is less than the threshold The temperature is determined to be stable; the trajectory offset rate is calculated by dividing the difference between the trajectories of two adjacent points by the sampling time interval, with the sampling frequency set to 100Hz, i.e., the time interval is... If the position difference between the trajectories of two points is Then the offset rate is Less than the set threshold If the trajectory continuity condition is met, and the system determines that all three conditions for this time slice are met, then this time slice is recorded as a "stable segment". If three consecutive time slices (such as...) occur, the system will not record this time slice as a "stable segment". If all three conditions are met, the system records the group as a "continuous stable segment group". The system continues to slide and count to the end of the sequence to obtain a complete continuous stable group sequence. For example, if the number of groups that meet the conditions is 5, the system will mark the result value as "5" and output the start and end indicators of the segment on the original time axis for subsequent matching operations.

[0136] The instruction generation submodule determines whether there are three adjacent time slices that simultaneously meet the three stability conditions based on the number of consecutive stable segments. If so, it extracts the corresponding deviation data and generates anomaly triggering parameters.

[0137] The instruction generation submodule obtains the number of consecutive stable segment groups output from the previous step and determines whether there are at least three adjacent time slices that simultaneously satisfy the three stability conditions. The system sets the judgment criterion as follows: three consecutive segment groups appear in the stable segment sequence and the time difference is less than 1. That is, satisfying the condition that the time span of each group is The start and end times of three adjacent groups should be as follows: , , During execution, the system first compares the stable fragment group indexes by group to obtain the starting time interval between adjacent groups. If the interval is... If the data is found to be in three consecutive groups, the system will determine that it is a "three-times-consecutive-times" data set and proceed with the next data extraction operation. The system will extract the abnormal deviation data from these three time periods, including the maximum vibration deviation value, the maximum temperature fluctuation value, and the maximum trajectory deviation rate value contained within that period. For example, the maximum acceleration recorded in each of the three time periods is... Temperature difference for trajectory speed is , the system records three values as trigger parameters in sequence, and then combines the three values to form an "abnormal trigger parameter set" containing time identifier , vibration trigger value , temperature trigger value , trajectory trigger value , and pushes the parameter set to the upper control system for linkage response trigger. In the stable segment group, the system cyclically traverses three adjacent structure combinations, and counts the number of time periods that meet the trigger condition. For example, if three continuous groups that meet the condition are detected, three abnormal trigger parameter sets are output, each set corresponding to a segment of stability mutation trigger state. The result will be used for further action judgment.

[0138] Specifically, as shown in Figure 2 , 6 , the parameter correction module includes:

[0139] An abnormality identification submodule obtains vibration amplitude data, temperature collection data, and trajectory offset vector value in the abnormal trigger parameter, calculates the average amplitude of the vibration time sequence, obtains the temperature change value per unit time to calculate the slope, calculates the offset direction according to the trajectory point displacement, and generates an abnormal feature judgment value.

[0140] When the abnormality identification submodule calls the abnormal trigger parameter, it reads three parameter items in sequence, which are vibration amplitude data, temperature collection data, and trajectory offset vector value. First, for the vibration amplitude data, the system reads the vibration acceleration value sequence corresponding to each time slice in the continuous stable segment group , where the corresponding values are , the sum is obtained by adding the three values in sequence , and then divided by the sample point number 3 to obtain the average value , thereby obtaining the average value of the vibration amplitude. Second, for the temperature collection data, the system calls the original temperature value data set collected in each group of time slices , and calculates the temperature change slope per unit time by dividing the temperature difference of each pair of adjacent points by the time interval. The time interval is fixed at , so the two slopes are and , and the average slope is . Then compare it with the system preset temperature slope reference value , calculate the relative offset ratio , and mark the temperature slope deviation amplitude as "0.5". The third item is to process the trajectory offset vector value. The system extracts the spatial three-axis trajectory position sequence , and constructs a trajectory vector set with each time slice as a unit, for example, time points , , corresponding to the coordinates of the trajectory points respectively , , , the system performs a vector difference operation on each pair of adjacent time slices, and calculates the direction vector as follows: the first segment is , the second segment is , and then by adding the two segments and performing normalization processing, the direction vector is calculated as:

[0141] ;

[0142] Subsequently, the unit direction vector is mapped to an azimuth encoding value, and the offset main direction is obtained by converting the spatial quadrant standard, for example, if the main direction corresponds to the first quadrant, the encoding is "D1", and the encoding result is used as the output of the trajectory offset direction judgment; after processing the three sub-items, the system generates an abnormal feature judgment value set, including the vibration amplitude average value , the temperature change slope average value , and the trajectory offset direction vector encoding "D1", which is used as an input item in the subsequent judgment and adjustment steps.

[0143] The safety interval adjustment submodule calls the vibration amplitude average value in the abnormal feature judgment value, sets the upper and lower limits of the initial safety interval of the insertion force, calculates the percentage correction amplitude, and generates the insertion force correction interval after adjusting the boundary value;

[0144] The safety interval adjustment submodule calls the vibration amplitude average value in the abnormal feature judgment value set , the system first sets the upper and lower limits of the initial safety interval of the insertion force, assuming that the initial setting is , , the called vibration amplitude average value needs to be linked with the insertion force fluctuation sensitivity coefficient to correct the boundary, and is set, then the percentage correction amplitude is represented by , where is the initial reference vibration value, and substituting the numerical value gives:

[0145] ;

[0146] The system determines that the current vibration value is in the high interval according to the offset amplitude, and performs bidirectional symmetric adjustment of the upper and lower limits of the insertion force interval, that is, the upper limit value is adjusted to:

[0147] ;

[0148] The lower limit value is adjusted to:

[0149] ;

[0150] If the vibration amplitude is lower than the reference value, reverse symmetric adjustment is performed, and in this case, since , the upper limit is reduced and the lower limit is raised to complete the compression processing; at the same time, it is necessary to determine whether the difference between the upper and lower limits after the correction is still greater than the minimum allowable interval width . After confirming that the conditions are met, the system generates a set of corrected plug-in force interval boundary values . If the width requirement is not met, the original set interval is retained and marked as an exception. The correction process also needs to record the adjustment time point to track the strategy change history, for example:

[0151] Table 4: Plug-in force safety interval adjustment table

[0152]

[0153] As shown in Table 4, after the vibration amplitude is high, the interval compression is triggered, and the compression result still meets the minimum allowable interval width requirement, so the system can normally enter the next correction process.

[0154] The correction set generation submodule calls the boundary values of the plug-in force correction interval, combines the temperature change slope and trajectory offset direction scalar in the exception feature judgment value, adjusts the stay duration based on the temperature slope, sets the path adjustment vector according to the offset direction, and generates a correction parameter set;

[0155] The correction set generation submodule first calls the boundary values of the plug-in force correction interval , combines the temperature change slope and the trajectory offset direction code "D1" to perform parameter calculation. The system first processes the temperature slope correction term, adjusts the stay duration based on the temperature slope, sets the temperature reference slope , adjustment coefficient , and the corrected stay duration value is:

[0156] ;

[0157] The original system set stay time is , and the corrected stay time is:

[0158] ;

[0159] Then, according to the trajectory offset direction, the path correction processing is performed. "D1" corresponds to the first quadrant direction, the system reads the corresponding direction vector from the direction code library as , and after unit normalization, it is . The path offset scale coefficient is set as , and the final path correction vector is:

[0160] ;

[0161] The system corrects the end effector trajectory control module with the path offset vector to form a path redirection parameter, while packaging the current dwell time, path correction amount, and correction insertion and extraction force interval into a final correction parameter set as follows: insertion and extraction force lower limit: 12.2664 N, insertion and extraction force upper limit: 17.7336 N, dwell time: 2.0 s, path offset vector: (0.4616, 0.4616, 0.4616) mm;

[0162] The system synchronously transmits the set to the control logic processing core to complete the dynamic adaptation of the subsequent insertion and extraction control strategy. The correction parameter set has derivation logic corresponding to the determination value in the previous step, and the parameter result has traceability through the formula calculation process, ensuring that the participating item value range is reasonable and falls within the normal physical interval, wherein the insertion and extraction force correction amount is reduced by 0.533 N from the original interval, the dwell time is shortened by 0.5 s, the path correction direction is clearly directed to the D1 direction vector, forming a set of dynamic control correction instruction set that can be used for actual execution.

[0163] Specifically, as shown in Figure 2 , 7 , the trajectory optimization execution module includes:

[0164] A trajectory reconstruction submodule that, based on the position offset and angle change value in the correction parameter set, uses a spline interpolation function to interpolate the three-dimensional coordinates of the insertion and extraction nodes in the original trajectory data, recalculates the time interval and position change between nodes, adjusts the smoothness of the trajectory curve, and obtains a trajectory smoothness and continuity value.

[0165] The trajectory reconstruction submodule processes based on the position offset and angle change value in the correction parameter set. The submodule first reads the three-dimensional coordinate point set of the insertion and extraction nodes in the original trajectory data of the end effector, such as the node set in the original trajectory, where each point extracts the position offset vector from the correction parameter set:

[0166] ;

[0167] and acts on each insertion and extraction node one by one, i.e., performs vector addition on the original coordinates, such as node , the corrected position is:

[0168] ;

[0169] The submodule will also perform an angle correction task, reading the angle offset value recorded by the sensor, and rotating and adjusting the insertion and extraction direction in the neighborhood around each three-dimensional point, such as rotating the effector direction vector around the z-axis to rotate the original direction vector Adjustment to , substitute , calculate , then perform spline interpolation on the node sequence after offset and rotation processing, smooth the nodes with cubic spline, for example, insert two points between nodes and , set to and respectively, construct the spline interpolation function with known node coordinates and timestamps , calculate the new interpolation point coordinates to make the trajectory change more smoothly in this time period, repeat this operation until the interpolation reconstruction between the inserted nodes is completed, then calculate the adjacent node time interval (where is the number of newly interpolated nodes) according to the reconstructed node timestamp set , combine the node coordinate transformation amplitude to obtain the change rate of each trajectory , and calculate the speed variance of the entire trajectory to measure the degree of change in trajectory smoothness, taking this value as the trajectory smoothness metric value, and comparing it with the reference continuity metric standard , if the current value is lower than the reference value, it is determined that the current trajectory meets the smoothness standard, if it is higher than the value, it is recorded as a trajectory segment to be adjusted, for example, a group of trajectory sample nodes are calculated as shown in the following table:

[0170] Table 5: Trajectory node interpolation and speed statistics table

[0171]

[0172] As shown in Table 5, the speed change after interpolation between 4 trajectory nodes is listed, and the trajectory smoothness metric value is further calculated by multiple speed calculations , the specific calculation is as follows:

[0173] ;

[0174] ;

[0175] The results show that the current trajectory smoothness metric value is higher than the set reference value , therefore the trajectory segment is recorded as a node segment that does not meet the smoothness standard after reconstruction, and enters the next stage of path adjustment submodule for sequential rearrangement and speed rhythm control processing.

[0176] The path adjustment submodule rearranges the path order between the trajectory nodes according to the trajectory smooth continuity metric value and the speed limit threshold in the correction parameter set, calls the spatial distance between the trajectory nodes and the contact order, adjusts the speed rhythm of the path based on the speed change, and generates a path order matching deviation rate;

[0177] The path adjustment submodule first receives the trajectory smooth continuity metric value passed by the trajectory reconstruction submodule and the speed limit threshold in the correction parameter set The module determines that the current smoothness does not meet the standard requirement, so the trajectory node rearrangement process is started, and the timestamp and the corresponding three-dimensional coordinate point in the node set are processed, first the spatial distance between adjacent nodes is calculated , then the corresponding time interval is calculated, and the average speed of multiple segments is obtained , for example, for a group of nodes , the calculation result is , then the difference between the speed and the preset limit threshold is calculated to obtain the speed threshold difference of each segment , which are , , , then the path spatial distance weight factor , the local path dynamic adjustment factor , and the contact order influence factor of the trajectory nodes are called, which are set to , , in the example, the path order matching deviation rate is calculated, and the formula is:

[0178] ;

[0179] Among them, represents the path order matching deviation rate, and represent the three-dimensional coordinates of the th and the th adjacent node, represents the time interval between the th node and the th node, represents the speed limit threshold, represents the spatial distance weight factor of the th node, represents the local path dynamic adjustment factor of the th node, represents the Contact sequence influence factor of a node, Indicates the sequence number of the trajectory node in the trajectory data, Indicates the total number of trajectory nodes.

[0180] Calculate the deviation weighting term of each item, respectively:

[0181] The first item: ;

[0182] The second item: ;

[0183] The third item: ;

[0184] After summation:

[0185] ;

[0186] The value represents the path sequence matching deviation rate is 1.5216, and the system judges whether the speed fluctuation and spatial arrangement between the current trajectory nodes exceed the expected running rhythm range, if the deviation rate is greater than the system set matching deviation reference upper limit , it is confirmed that there is a speed rhythm imbalance problem in the current path, and the path node sequence needs to be rearranged. The rearrangement operation is sorted according to the priority of the combination of the speed and contact sequence factor of the node, for example, taking the weighted sorting score as the standard, the node corresponding weight is 0.72, 0.792, 1.2 respectively, and the trajectory node sequence is adjusted from high to low according to the weight as the third item, the second item, the first item, and then the time interval is rearranged to match the speed and spatial distance requirements. The system outputs the new path node arrangement sequence and the corrected speed distribution data, and recalculates the new path sequence matching deviation rate to confirm that it is lower than the reference value before entering the next stage of processing flow.

[0187] Error correction submodule, according to the path sequence matching deviation rate and the interpolated trajectory data, according to the current position deviation and speed trajectory offset data of the actuator, the three-dimensional trajectory offset degree is compared, the trajectory node which exceeds the speed change reference value is screened, and the optimized trajectory execution instruction is generated;

[0188] The error correction submodule is based on the path sequence matching deviation rate and the trajectory data after the spline interpolation processing, first call the current actual pose information of the actuator, that is, at the current time point collected by the system in real time, get the current position coordinates and the current speed vector of the end effector, then read the theoretical position of the corresponding time point in the interpolation trajectory:

[0189] ;

[0190] with the theoretical speed vector , perform a position deviation calculation operation , recalculate the speed offset , for the spatial position deviation vector modulus value , combine the system-set three-dimensional trajectory offset reference value , perform an offset difference judgment, i.e., perform a comparison operation is true, so the node is marked as an error overrun node, and then the system sequentially traverses and processes the nodes in the entire trajectory, compares the actual pose of each node with the theoretical pose to perform the above position and speed difference judgment, and filters out a node set that satisfies the simultaneous offset exceeding the threshold value In the current instance, the trajectory node set is set as to , wherein the position deviations are {0.42, 0.28, 0.31, 0.56, 0.48, 0.60, 0.49, 0.52, 0.29, 0.26} (unit: mm), and the system performs a batch judgment operation , wherein the 4th, 6th, and 8th nodes satisfy the overrun condition, forming an error node subset , which is input into the error correction module, and then an instruction optimization operation is performed to calculate a compensation vector for the position and speed error of each node. The original theoretical position of the 6th node is set as , the current position is , and the position correction vector is . The correction vector is applied to the trajectory point to generate a corrected path point, and a synchronous adjustment of the speed instruction is performed according to the speed offset vector , e.g., if the speed deviation is , then a speed compensation adjustment is separately performed for each direction, and the optimized instruction set is in the form of:

[0191] ;

[0192] , wherein the coordinates are the corrected target point coordinates, and the speed is the difference compensation adjusted value. Finally, the instructions are output to the actuator control system in the form of an optimized sequence for replacing the execution instructions of the corresponding nodes in the original trajectory.

[0193] Please refer to Figure 8 The full-automatic plug-in test machine control method is based on the above-mentioned full-automatic plug-in test machine control system and includes the following steps:

[0194] S1: Collect acceleration, displacement change, temperature gradient, three-dimensional path coordinates, plug-in force and contact duration time series data through the sensor array, align vibration, temperature, trajectory, force data by timestamp, and generate a synchronized data set;

[0195] S2: Call the synchronized data set, use fast Fourier transform to process the vibration amplitude sequence to extract frequency characteristics, calculate the period amplitude difference and temperature change and perform comparison, combine the trajectory offset amount comparison, and generate a three-dimensional deviation set;

[0196] S3: Receive the three-dimensional deviation set, scan and identify the vibration super-stable band, temperature super-thermal stable range, and trajectory offset super-continuous range time slice, and send a stop command if the three time slices meet the conditions, and generate an abnormal trigger parameter;

[0197] S4: Receive the abnormal trigger parameter, adjust the plug-in force safety interval based on the average vibration amplitude, correct the stay time according to the temperature change slope, calculate the path fine-tuning vector according to the trajectory offset direction, and generate a correction parameter set;

[0198] S5: Call the correction parameter set, use a spline interpolation function to perform smoothing operation on the three-dimensional path, re-plan the plug-in path sequence, call the speed curve to adjust the speed parameter of the operation, rearrange the contact timing to get the corresponding plug-in path, check the three-dimensional deviation state of the corrected trajectory, and output the optimized trajectory execution instruction.

[0199] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A control system for a full automatic plug test machine, characterized in that, The system comprises: A multi-dimensional data acquisition module acquires time series data of acceleration, displacement change, temperature gradient, three-dimensional path coordinates, plug-in force and contact duration through a sensor array, aligns vibration, temperature, trajectory and force data according to timestamps, generates a synchronous data set, and delivers it to an abnormal state identification module; The abnormal state identification module receives the synchronous data set, uses fast Fourier transform to process vibration amplitude sequence to extract frequency characteristics, calculates and compares period amplitude difference and temperature change, combines trajectory offset comparison, generates a three-dimensional deviation set, and delivers it to an action termination judgment module; The action termination judgment module receives the three-dimensional deviation set, scans and identifies vibration super-stable band, temperature super-thermal stable range and trajectory offset super-continuous range time slice, sends a stop command if the three time slices meet the conditions, generates an abnormal trigger parameter, and delivers it to a parameter correction module; The parameter correction module receives the abnormal trigger parameter, adjusts the plug-in force safety interval based on the average vibration amplitude, corrects the stay duration according to the temperature change slope, calculates the path fine-tuning vector according to the trajectory offset direction, generates a corrected parameter set, and delivers it to a trajectory optimization execution module.

2. The control system for a full automatic plug test machine according to claim 1, wherein, The synchronous data set includes data time alignment results, signal synchronization accuracy indicators and time series integrity verification, the three-dimensional deviation set includes frequency spectrum characteristic difference values, temperature gradient change amplitudes and spatial trajectory deviation amounts, the abnormal trigger parameter includes continuous abnormal state judgment identifiers, device protection start signals and abnormal action control codes, and the corrected parameter set includes plug-in force safety threshold intervals, temperature change rate correction coefficients and trajectory offset vector adjustment amounts.

3. The control system for a full automatic plug test machine according to claim 1, wherein, The multi-dimensional data acquisition module comprises: A vibration temperature acquisition sub-module acquires time series data of acceleration, displacement change, temperature gradient, three-dimensional path coordinates, plug-in force and contact duration through a sensor array, aligns acceleration and temperature gradient data based on timestamps, calculates the correlation between acceleration and temperature change, and generates a warming correlation coefficient; A trajectory data generation sub-module, based on the warming correlation coefficient, calls displacement change data, calculates displacement increments and direction vectors for each time period, applies coordinate transformation, normalizes displacement change amounts according to time series, and generates three-dimensional path change amounts; A mechanical characteristic alignment sub-module, according to the three-dimensional path change amounts, calls plug-in force and contact duration data, calculates the increment of plug-in force change and the relative change of contact duration, aligns data based on the data relationship between plug-in force and contact duration, and generates a synchronous data set.

4. The control system for a full automatic plug test machine according to claim 3, wherein, The warming correlation coefficient refers to a numerical parameter calculated by correlation analysis algorithm based on acceleration data and temperature gradient data of the same timestamp; The normalization processing refers to using the range normalization method to linearly transform the original three-dimensional path change amounts to a unified interval; The three-dimensional path change amounts refer to vectors composed of displacement increments in X, Y and Z three-axis directions calculated according to displacement change data at different time intervals.

5. The control system for a full automatic plug test machine according to claim 3, wherein, The abnormal state identification module comprises: The frequency feature extraction submodule extracts each sampling point value of the vibration amplitude sequence based on the synchronization data set, establishes an amplitude array in time sequence, calculates a frequency corresponding amplitude value sequence using fast Fourier transform, aggregates the amplitude value proportion according to frequency distribution, and generates a frequency amplitude value distribution coefficient; The amplitude-temperature difference calculation submodule calculates the continuous period amplitude value difference and temperature difference according to the frequency amplitude value distribution coefficient and the temperature sequence at the same time in the synchronization data set, arranges the values correspondingly, integrates them in time sequence, and obtains the period amplitude difference-temperature difference ratio; The offset ratio generation submodule calls the period amplitude difference-temperature difference ratio and the trajectory offset sequence, constructs the offset trend and ratio curve for the same time period, extracts the corresponding amplitude sequence, performs amplitude normalization matching according to the time sequence, maps it to a three-dimensional coordinate, aggregates the coordinate points, and generates a three-dimensional deviation set.

6. The control system for a full automatic plug test machine according to claim 5, wherein, The action termination judgment module includes: The deviation recognition submodule obtains the acceleration deviation, temperature deviation, and trajectory offset value in the three-dimensional deviation set, judges whether each time slice exceeds the corresponding threshold value according to the vibration recognition threshold value, temperature recognition threshold value, and trajectory offset threshold value, extracts abnormal deviation time slice data, and generates an abnormal deviation amplitude interval; The stable region scanning submodule calls the vibration amplitude value, temperature change range, and trajectory offset rate in the corresponding time slice based on the abnormal deviation amplitude interval, filters the time slices that meet the stable condition according to the vibration stable judgment interval, temperature stable interval, and trajectory continuous range, counts the number of continuously appearing groups, and generates the number of continuous stable segments. The instruction generation submodule judges whether there is a time period in which three adjacent time slices simultaneously meet the three stable conditions according to the number of continuous stable segments, extracts the corresponding deviation data if the condition is met, and generates an abnormal trigger parameter.

7. The control system for a full automatic plug test machine according to claim 6, wherein, The parameter correction module includes: The abnormal recognition submodule obtains the vibration amplitude data, temperature acquisition data, and trajectory offset vector value in the abnormal trigger parameter, calculates the average amplitude of the vibration time sequence, obtains the temperature change value per unit time to calculate the slope, calculates the offset direction according to the trajectory point displacement, and generates an abnormal feature judgment value. The safe interval adjustment submodule calls the vibration amplitude average value in the abnormal feature judgment value, sets the upper and lower limits of the initial safe interval of the insertion force, calculates the percentage correction amplitude, adjusts the boundary value, and generates a corrected insertion force interval. The correction set generation submodule calls the boundary value of the corrected insertion force interval, combines the temperature change slope and trajectory offset direction scalar in the abnormal feature judgment value, adjusts the stay time according to the temperature slope, sets the path adjustment vector according to the offset direction, and generates a corrected parameter set.

8. The control system for a full automatic plug test machine according to claim 1, wherein, The trajectory optimization execution module receives the corrected parameter set, smoothes the trajectory data using a spline interpolation function, re-plans the insertion path sequence, adjusts the speed curve and contact time sequence, executes the new trajectory three-dimensional deviation state, and generates an optimized trajectory execution instruction. The optimized trajectory execution instruction includes a spline interpolation curve parameter, a speed curve shape adjustment parameter, and a contact time sequence optimization arrangement.

9. The control system for a full automatic plug test machine according to claim 8, wherein, The trajectory optimization execution module includes: The trajectory reconstruction submodule uses a spline interpolation function to perform interpolation processing on the three-dimensional coordinates of the plug-in nodes in the original trajectory data based on the position offset and the angle change value in the set of correction parameters, recalculates the time interval and position change between the nodes, adjusts the smoothness of the trajectory curve, and obtains a trajectory smoothness and continuity metric value; The path adjustment submodule rearranges the path order between the trajectory nodes according to the trajectory smoothness and continuity metric value and the speed limit threshold in the set of correction parameters, calls the spatial distance and contact order between the trajectory nodes, adjusts the speed rhythm of the path based on the speed change, and generates a path order matching deviation rate; The error correction submodule compares the three-dimensional trajectory deviation degree according to the path order matching deviation rate and the interpolated trajectory data, according to the current position deviation and speed trajectory offset data of the actuator, filters out the trajectory nodes that exceed the speed change reference value, and generates an optimized trajectory execution instruction.

10. A control method for a full automatic plug test machine, characterized in that, The full-automatic plug-in test machine control system execution according to any one of claims 1-9, comprising the following steps: S1: Collecting acceleration, displacement change, temperature gradient, three-dimensional path coordinates, plug-in force and contact duration time series data through a sensor array, aligning vibration, temperature, trajectory, force data by timestamp, and generating a synchronous data set; S2: Calling the synchronous data set, using fast Fourier transform to process vibration amplitude sequence to extract frequency characteristics, calculating and comparing period amplitude difference and temperature change, combining trajectory offset comparison, and generating a three-dimensional deviation set; S3: Receiving the three-dimensional deviation set, scanning and identifying vibration super-stable band, temperature super-thermal stable range, trajectory offset super-continuous range time slice, sending a stop instruction if the continuous three time slices meet the conditions, and generating an abnormal trigger parameter; S4: Receiving the abnormal trigger parameter, adjusting the plug-in force safety interval based on the average vibration amplitude, correcting the stay time according to the temperature change slope, calculating the path fine-tuning vector according to the trajectory offset direction, and generating a set of correction parameters; S5: Calling the set of correction parameters, performing smoothing operation on the three-dimensional path using a spline interpolation function, re-planning the plug-in path sequence, calling the speed curve to adjust the speed parameter, rearranging the contact time sequence to obtain the corresponding plug-in path, verifying the three-dimensional deviation state of the corrected trajectory, and outputting an optimized trajectory execution instruction.

Images