Intelligent assembly optimization platform and method for wire harness

Abstract

The invention belongs to the technical field of intelligence, and discloses a wiring harness intelligent assembly optimization platform and method. Comprising the following steps: storing perceived data; performing optimization through a group search optimization mechanism based on the terminal process data and the assembly equipment state data to obtain optimal process data; wherein in the group search optimization mechanism, the fitness value of the position of each individual is calculated by using a staged mixed strategy method; a double repair boundary optimization method is used to repair the cross-border feature variables in the updated individual positions; based on the optimal process data, controlling the wiring harness and the terminal to be assembled; after assembly, predicting electrical connectivity data; performing correction to obtain dynamic quality judgment data resisting environmental interference; whether assembly is qualified or not is judged based on the dynamic quality judgment data resisting environmental interference; if not, reassembling is carried out; the wire harness assembling precision and efficiency are effectively improved, the production cost is reduced, and the production period is shortened.

Description

Technical Field

[0001] The present invention relates to the field of intelligent technologies, and particularly to an intelligent assembly optimization platform and method for wire harnesses. Background Art

[0002] Existing intelligent assembly optimization systems for wire harnesses have many problems. First, process data is set for installation through empirical methods or existing optimization algorithms. However, empirical methods are highly subjective, difficult to adapt to complex process changes, and lack the ability for continuous optimization. Existing optimization algorithms are prone to falling into local optima, have a slow convergence speed, and handle outliers and parameter out-of-bounds poorly, resulting in unstable process data and large fluctuations in assembly quality.

[0003] In the existing assembly process, there is a lack of real-time detection, resulting in problems not being discovered in a timely manner. They are often only discovered in subsequent inspection links, which may lead to unqualified products flowing into the market, increasing production costs, delaying delivery times, and reducing overall production efficiency.

[0004] Existing detection technologies often ignore the interference of some external factors on the assembly process. These factors are not monitored or corrected in a timely manner, resulting in unstable assembly quality and potential failure risks.

[0005] In view of this, the present invention proposes an intelligent assembly optimization platform and method for wire harnesses to solve the above problems. Summary of the Invention

[0006] In order to overcome the above defects of the prior art and to achieve the above object, the present invention provides the following technical solution: An intelligent assembly optimization platform for wire harnesses, comprising:

[0007] A multi-source data sensing subsystem: capturing terminal process data, assembly equipment status data, electrical performance data, environmental data, vibration data, and electromagnetic interference data in real time;

[0008] A process parameter optimization engine subsystem: relying on terminal process data and assembly equipment status data, generating optimal process data through a population search optimization mechanism; in the population search optimization mechanism, a phased hybrid strategy is used to calculate the fitness value, and the out-of-bounds variables are corrected by combining the leading individual position update with the dual repair boundary optimization method, and the optimal process data is output;

[0009] Among them, the phased hybrid strategy for calculating the fitness value is to calculate the deviation value and stability evaluation value in an individual; according to the number of executions, the fitness value of each individual is calculated using the deviation value and stability evaluation value;

[0010] Among them, in the dual repair boundary optimization method, the neighborhood search repair method is used to repair the updated individual position; then the boundary dynamic projection method is used to repair the updated individual position after repair;

[0011] Precision Assembly Execution Control Subsystem: Input the optimal process data into the assembly equipment control system to drive the automated assembly of wire harnesses and terminals.

[0012] Anti-Environmental Interference Dynamic Quality Assessment Subsystem: For the assembled wire harness, based on electrical performance data and environmental disturbance data, construct an electrical connectivity prediction model to predict electrical connectivity data; fuse vibration data and electromagnetic interference data to construct physical field feedback signal data, and correct the predicted electrical connectivity data to form dynamic quality judgment data resistant to environmental interference.

[0013] Real-Time Monitoring Subsystem: Based on the dynamic quality judgment data resistant to environmental interference, determine whether the assembly is qualified; if the dynamic quality judgment data resistant to environmental interference is good, the assembly is qualified; if the dynamic quality judgment data resistant to environmental interference is poor, the assembly is unqualified, and dynamically adjust the dual repair boundary optimization method in the population search mechanism to regenerate new optimal process data, and perform assembly according to the new optimal process data.

[0014] Furthermore, the terminal process data includes crimping force data, insertion force data, and extraction force data.

[0015] The assembly equipment status data includes equipment operating speed data and equipment operating temperature data.

[0016] The electrical performance data includes voltage data, current data, contact resistance data, conduction resistance data, and insulation resistance data.

[0017] The environmental data includes environmental temperature data and environmental humidity data.

[0018] Furthermore, the specific method for generating the optimal process data through the population search optimization mechanism relying on the terminal process data and the assembly equipment status data includes:

[0019] Step A1: Take a set of terminal process data and assembly equipment status data as an individual; set the number of individuals; set the maximum number of executions.

[0020] Step A2: Set a standard range for the terminal process data and the state data of the transfer equipment, and randomly initialize the data values within the standard range as the position of the individual.

[0021] Step A3: Use the phased hybrid strategy method to calculate the fitness value of the position of each individual.

[0022] Step A4: During the execution process, select the leading individuals and update their positions; the leading individuals include Individual 1, Individual 2, and Individual 3. Individual 1 is the optimal individual with the best fitness, Individual 2 is the second-best individual with the second-best fitness, and Individual 3 is the third-best individual with the third-best fitness;

[0023] Step A5: Update the positions according to the leading individuals 1, 2, and 3. The formula is: Where, represents the position of the i-th updated individual, X i represents the position of the i-th individual. A1, A2, and A3 respectively represent the random coefficients of Individuals 1, 2, and 3. The formula is: A1 = 2 * a * r1 - a, A2 = 2 * a * r2 - a, A3 = 2 * a * r3 - a. Where, r1, r2, and r3 represent random numbers with a value range from 0 to 1, and a represents a constant that gradually decreases with the number of executions. D1, D2, and D3 respectively represent the distances between the current individual position and Individuals 1, 2, and 3. The formula is: D1 = |C * X1 - X i |, D2 = |C * X2 - X i |, D3 = |C * X3 - X i |. Where, X1, X2, and X3 are the positions of Individuals 1, 2, and 3 respectively, and C represents a random coefficient. The formula is: C = 2 * R, where R is a random number with a value range from 0 to 1; Where, represents the first characteristic variable in the position of the i-th updated individual, represents the second characteristic variable in the position of the i-th updated individual, represents the d-th characteristic variable in the position of the i-th updated individual, and d is the total number of characteristic variables;

[0024] Step A6: After the position update is completed, use the double repair boundary optimization method to repair the out-of-bounds characteristic variables in the updated individual positions;

[0025] Step A7: Repeat Steps A4 to A6 until the set maximum number of executions is reached, and output a set of optimal process data, that is, the optimal terminal process data and assembly equipment status data.

[0026] Furthermore, the specific method of using the phased hybrid strategy method to calculate the fitness value of each individual's position includes:

[0027] For the terminal process data of an individual, calculate the absolute deviation amounts of the crimping force, insertion force, and extraction force in the terminal process data from the ideal crimping force, ideal insertion force, and ideal extraction force, and obtain the deviation degree value based on the ratios of the deviation amounts to the standard ranges of the crimping force, insertion force, and extraction force; among them, the ideal crimping force, ideal insertion force, and ideal extraction force are obtained through the connector industry standard;

[0028] For the assembly equipment status data of an individual, extract the fluctuation standard value of the equipment operating speed data, and calculate the fluctuation ratio of the equipment operating speed data with the reference value of the equipment operating speed data; calculate the temperature deviation ratio of the equipment operating temperature data by taking the absolute difference between the equipment operating temperature data and the reference operating temperature data and dividing it by the reference operating temperature data; calculate the stability evaluation value based on the fluctuation ratio of the equipment operating speed data and the temperature deviation ratio of the equipment operating temperature data;

[0029] When the number of executions is the first 30%, mainly based on the deviation degree value and supplemented by the stability evaluation value, use the dynamic attenuation function to adjust the weights of the deviation degree value and the stability evaluation value, and calculate the fitness value of the individual;

[0030] When the number of executions is the last 70%, mainly based on the stability evaluation value and supplemented by the deviation degree value, use the dynamic attenuation function to adjust the weights of the stability evaluation value and the deviation degree value, and calculate the fitness value of the individual.

[0031] Further, the specific method for repairing the out-of-bounds characteristic variables in the updated individual position using the double repair boundary optimization method includes:

[0032] Use the np.max method and np.min method to obtain the maximum and minimum values of the terminal process data and the assembly equipment status data in the individual position, that is, the maximum and minimum values of the characteristic variables, and use the maximum value as the upper boundary and the minimum value as the lower boundary;

[0033] If Or Then it means that the j-th characteristic variable of the i-th updated individual position is out of bounds, where Lb j Represents the upper boundary of the j-th characteristic variable, and Ub j Represents the lower boundary of the j-th characteristic variable;

[0034] For the terminal process data and the assembly equipment status data in the out-of-bounds updated individual position, use the neighborhood search repair method for repair, and calculate the neighborhood radius. The formula is: Among them, rd t,j Is the neighborhood radius of the j-th characteristic variable in the t-th execution. As the execution progresses, the neighborhood radius gradually decreases. t is the current execution number, and Tmax is the maximum number of executions, and γ is the shrinkage rate parameter;

[0035] For the out-of-bounds terminal process data and assembly equipment status data in the updated individual position, use the terminal process data and assembly equipment status data in the position of individual 1 during the current execution and the neighborhood radius to perform repair, and obtain the repaired updated individual position. The formula is: Among them, α t,j represents the j-th characteristic variable in the position of individual 1 at the t-th execution, and rand(-rd t,j , rd t,j ) represents a random number randomly generated in the interval [-rd t,j , rd t,j . represents the repair value of the j-th characteristic variable in the i-th updated individual position; for the repair value, further judge whether it is out of bounds. If represents exceeding the upper limit, then define the upper limit Lb j as the repair value, or represents exceeding the lower limit, then define the lower limit Ub j as the repair value.

[0036] Furthermore, the specific steps of inputting the optimal process data into the assembly equipment control system to drive the automatic assembly of the wire harness and the terminal include:

[0037] Input the optimal process data, that is, the optimal terminal process data and assembly equipment status data, into the assembly equipment control subsystem. The system executes commands according to the input optimal process data to assemble and connect the wire harness and the terminal to complete the assembly.

[0038] Furthermore, the specific method of constructing an electrical connectivity prediction model based on electrical performance data and environmental disturbance data to predict electrical connectivity data includes:

[0039] Perform missing value, outlier, and normalization processing on the electrical performance data and environmental data to obtain a preprocessed data set;

[0040] Input M_M samples, and each group of samples includes a preprocessed data set and the corresponding electrical connectivity data;

[0041] Use the random method to initialize the hyperparameters, weight vector, and bias term. The hyperparameters include the penalty coefficient C, epsilon parameter, and gamma parameter; set the number of executions;

[0042] Calculate the initial loss function value of the model;

[0043] Use the gradient descent method to update the weight vector and bias term during each execution, and calculate the new loss function during the update process.

[0044] When the set number of executions is reached, the execution stops and the predicted electrical connectivity data is output.

[0045] Furthermore, the method for obtaining the dynamic quality judgment data resistant to environmental interference includes:

[0046] Based on the vibration data and electromagnetic interference data, the linear regression method is used to construct the physical field feedback signal data, WL = 1 - a1*V - a2*EMI; where, a1 represents the influence coefficient of the vibration data on the electrical connectivity, that is, the vibration influence degree coefficient, V represents the vibration data, a2 represents the influence coefficient of the electromagnetic interference data on the electrical connectivity, that is, the electromagnetic interference influence degree coefficient, EMI represents the electromagnetic interference data, and WL represents the physical field feedback signal data;

[0047] Based on the predicted electrical connectivity data, the physical field feedback signal data is used for correction to form the dynamic quality judgment data resistant to environmental interference. The formula is: DC fin = DC pre *WL, where, DC fin is the dynamic quality judgment data formed to resist environmental interference, and DC pre is the predicted electrical connectivity data.

[0048] Furthermore, the specific method for judging whether the assembly is qualified based on the dynamic quality judgment data resistant to environmental interference includes:

[0049] When the dynamic quality judgment data formed to resist environmental interference is greater than 0.5, it indicates good electrical connectivity, and the assembly is qualified; when the dynamic quality judgment data formed to resist environmental interference is less than or equal to 0.5, it indicates poor electrical connectivity, and the assembly is unqualified.

[0050] A method for optimizing the intelligent assembly of wire harnesses, which is implemented by applying to the intelligent assembly optimization platform of the wire harness, includes:

[0051] Step SS1: Real-time capture terminal process data, assembly equipment status data, electrical performance data, environmental data, vibration data, and electromagnetic interference data;

[0052] Step SS2: Relying on the terminal process data and assembly equipment status data, generate the optimal process data through the group search optimization mechanism; in the group search optimization mechanism, the fitness value is calculated using the phased hybrid strategy, and the out-of-bounds variables are corrected by combining the leader individual position update with the double repair boundary optimization method, and the optimal process data is output;

[0053] Among them, calculating the fitness value using the phased hybrid strategy is to calculate the deviation value and stability evaluation value in the individual; according to the number of executions, the fitness value of each individual is calculated using the deviation value and stability evaluation value;

[0054] Among them, in the double repair boundary optimization method, the neighborhood search repair method is used to repair the updated individual position; then the boundary dynamic projection method is used to repair the updated individual position after repair.

[0055] Step SS3: Input the optimal process data into the control system of the assembly equipment to drive the automatic assembly of the wire harness and terminals.

[0056] Step SS4: For the assembled wire harness, based on the electrical performance data and environmental disturbance data, construct an electrical connectivity prediction model to predict the electrical connectivity data; fuse the vibration data and electromagnetic interference data to construct the physical field feedback signal data, and correct the predicted electrical connectivity data to form the dynamic quality judgment data against environmental interference.

[0057] Step SS5: Judge whether the assembly is qualified based on the dynamic quality judgment data against environmental interference; if the dynamic quality judgment data against environmental interference is good, the assembly is qualified; if the dynamic quality judgment data against environmental interference is poor, the assembly is unqualified, and the double repair boundary optimization method in the population search mechanism is dynamically adjusted to regenerate new optimal process data, and the assembly is carried out according to the new optimal process data.

[0058] The technical effects and advantages of the intelligent assembly optimization platform and method for wire harness of the present invention:

[0059] The present invention uses the multi-source data sensing subsystem to collect and monitor the terminal process data, assembly equipment status data, electrical performance data, environmental data, vibration data and electromagnetic interference data in real time, and the system can perform real-time perception and feedback on each link in the assembly process.

[0060] The process parameter optimization engine is based on the population search optimization mechanism, and through the phased hybrid strategy method and the double repair boundary optimization method, accurately optimizes each process parameter, such as the pressing force, insertion force and pulling-out force, etc., to ensure the best efficiency and stability of the assembly process, thereby significantly improving the assembly accuracy and reducing the assembly problems caused by improper process settings.

[0061] The precision assembly execution controller accurately controls the assembly process of the wire harness and terminals according to the optimal process data to ensure the high precision and high stability of the assembly.

[0062] By combining the electrical performance data and environmental data for electrical connectivity prediction, and further using the vibration data and electromagnetic interference data to correct the prediction results, the assembly quality can be accurately judged whether it is qualified. This process can not only predict the assembly result in advance, but also adjust according to external factors, avoiding the neglect of external interference factors by the detection method, and improving the reliability and long-term performance of the assembly. Description of the Drawings

[0063] Figure 1 Schematic diagram of an intelligent assembly optimization platform for a wire harness of the present invention;

[0064] Figure 2 Schematic diagram of the process parameter optimization engine subsystem of the present invention;

[0065] Figure 3 Schematic diagram of a method for optimizing intelligent assembly of a wire harness of the present invention. Detailed implementation manners

[0066] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0067] Embodiment 1

[0068] Please refer to Figure 1 and Figure 2 As shown, an intelligent assembly optimization platform for a wire harness in this embodiment includes:

[0069] Multi-source data sensing subsystem: Real-time capture of terminal process data, assembly equipment status data, electrical performance data, environmental data, vibration data, and electromagnetic interference data;

[0070] Process parameter optimization engine subsystem: Relying on terminal process data and assembly equipment status data, generating optimal process data through a population search optimization mechanism; in the population search optimization mechanism, using a phased hybrid strategy to calculate the fitness value, correcting out-of-bounds variables by combining the leading individual position update with a dual repair boundary optimization method, and executing the output of the optimal process data;

[0071] Among them, the phased hybrid strategy for calculating the fitness value is to calculate the deviation value and stability evaluation value in the individual; according to the number of executions, calculate the fitness value of each individual using the deviation value and stability evaluation value;

[0072] Among them, in the dual repair boundary optimization method, the neighborhood search repair method is used to repair the updated individual position; then the boundary dynamic projection method is used to repair the updated individual position after repair;

[0073] Precision assembly execution control subsystem: Input the optimal process data into the assembly equipment control system to drive the automated assembly of the wire harness and terminals;

[0074] Anti-Environmental-Interference Dynamic Quality Assessment Subsystem: For the assembled wire harness, based on electrical performance data and environmental disturbance data, construct an electrical connectivity prediction model to predict electrical connectivity data; fuse vibration data and electromagnetic interference data to construct physical field feedback signal data, and correct the predicted electrical connectivity data to form anti-environmental-interference dynamic quality judgment data;

[0075] Real-time Monitoring Subsystem: Judge whether the assembly is qualified based on the anti-environmental-interference dynamic quality judgment data; if the anti-environmental-interference dynamic quality judgment data is good, the assembly is qualified; if the anti-environmental-interference dynamic quality judgment data is poor, the assembly is unqualified, and dynamically adjust the double repair boundary optimization method in the population search mechanism to regenerate new optimal process data, and perform assembly according to the new optimal process data.

[0076] Terminal process data includes crimping force data, insertion force data, and extraction force data; install force sensors, thrust sensors, and tension sensors on the assembly equipment to collect crimping force data, insertion force data, and extraction force data;

[0077] Assembly equipment status data includes equipment running speed data and equipment running temperature data; use an optical encoder and a thermistor to collect equipment running data and equipment running temperature data;

[0078] Electrical performance data includes voltage data, current data, contact resistance data, conduction resistance data, and insulation resistance data; collect voltage data through a voltage sensor, collect current data through a current sensor, and collect contact resistance data, conduction resistance data, and insulation resistance data through a micro-ohmmeter;

[0079] Environmental data includes environmental temperature data and environmental humidity data; collect temperature data through a temperature sensor and collect humidity data through a humidity sensor;

[0080] Vibration data is collected through an acceleration sensor;

[0081] Electromagnetic interference data is collected through an electromagnetic field sensor;

[0082] The specific methods of the embedded sensing unit include:

[0083] After the sensors are installed, select STM32 as the microcontroller. The controller needs to have sufficient processing power to support data collection, signal processing, and real-time decision-making logic of various sensors;

[0084] Connect the sensors to the STM32: For sensors that output analog signals (such as temperature sensors, force sensors, etc.), connect them to the analog input interface (ADC) of the controller; for sensors that output digital signals (such as optical encoders, current sensors, etc.), connect them to the digital interfaces (I2C, SPI, UART) of the controller.

[0085] If the sensor outputs an analog signal, use an amplifier and a filter to condition the signal to eliminate noise and enhance signal quality; input the conditioned analog signal into the ADC module for digital processing; for digital signals, directly collect data through the I2C / SPI interface of the controller without additional conditioning circuits.

[0086] Write or call relevant sensor driver programs to support data reading and transmission. For analog sensors, the driver program needs to handle ADC sampling; for digital sensors, the driver program needs to handle data parsing and transmission.

[0087] Define the data acquisition period, for example, collect data once per second, and ensure that the embedded system collects sensor data on time.

[0088] Store the collected data in the internal memory of the embedded device or an SD card, or directly transmit it to the host computer for further processing.

[0089] Existing assembly processes often rely on rules of thumb or standardized process parameters and are difficult to flexibly meet complex assembly requirements; these process parameters usually cannot be adjusted according to specific situations, and it is easy to have excessive or insufficient assembly force, resulting in unstable assembly quality.

[0090] The specific ways to generate optimal process data by relying on terminal process data and assembly equipment status data through a population search optimization mechanism include:

[0091] Step A1: Take a set of terminal process data and assembly equipment status data as an individual; set the number of individuals; set the maximum number of executions.

[0092] Step A2: Set a standard range for the terminal process data and the status data of the assembly equipment, and randomly initialize the data values within the standard range as the positions of the individuals.

[0093] Step A3: Use the phased hybrid strategy method to calculate the fitness value of the position of each individual.

[0094] Step A4: During the execution process, select the leading individuals for position update; the leading individuals include individual 1, individual 2, and individual 3. Individual 1 is the optimal individual with the best fitness, individual 2 is the second-best individual with the second-best fitness, and individual 3 is the third-best individual with the third-best fitness.

[0095] Step A5: Update the position according to the leading individuals 1, 2, and 3. The formula is as follows: Wherein, represents the position of the i-th updated individual, X i represents the position of the i-th individual. A1, A2, and A3 respectively represent the random coefficients of individuals 1, 2, and 3. The formula is: A1 = 2 * a * r1 - a, A2 = 2 * a * r2 - a, A3 = 2 * a * r3 - a. Wherein, r1, r2, and r3 represent random numbers, and the value range is from 0 to 1. a represents a constant that gradually decreases with the number of executions. D1, D2, and D3 respectively represent the distances between the current individual position and individuals 1, 2, and 3. The formula is: D1 = |C * X1 - X i |, D2 = |C * X2 - X i |, D3 = |C * X3 - X i |. Wherein, X1, X2, and X3 are respectively the positions of individuals 1, 2, and 3. C represents a random coefficient. The formula is: C = 2 * R. Wherein, R is a random number, and the value range is from 0 to 1; Wherein, represents the first feature variable in the position of the i-th updated individual, represents the second feature variable in the position of the i-th updated individual, represents the d-th feature variable in the position of the i-th updated individual, and d is the total number of feature variables;

[0096] Step A6: After the position update is completed, use the double repair boundary optimization method to repair the out-of-bounds feature variables in the updated individual position;

[0097] Step A7: Repeat Steps A4 to A6 until the set maximum number of executions is reached, and output a set of optimal process data, that is, the optimal terminal process data and the assembly equipment status data;

[0098] The existing intelligent assembly optimization system for wire harnesses relies on the empirical method to set process data, lacks data learning and optimization capabilities, resulting in large fluctuations in assembly quality, low production efficiency, and inability to continuously improve. In contrast, through the group search optimization mechanism, combined with terminal process data and equipment status data, intelligent optimization, adaptive adjustment, and execution optimization are realized, accurately matching the optimal process parameters, not only improving the assembly quality and production efficiency, but also dynamically adjusting the process to adapt to different equipment states, ensuring continuous optimization and improvement, and thoroughly making up for the deficiencies of the existing methods.

[0099] Existing fitness calculation methods use weighted average or squared error methods, which ignore the requirements of different execution stages. In the initial stage of optimization, excessive focus on squared error may lead to premature convergence of the algorithm, missing the opportunity to widely explore the possible solution space. In the later stage, simple weighted average methods cannot refine the optimization, easily leading to local optima and thus unable to obtain the global optimal solution. Such calculation methods are prone to falling into local minima, resulting in insufficient accuracy of calculation results and accuracy of fitness.

[0100] Existing fitness calculation methods usually rely on fixed weight allocation and single-dimensional indicators, making it difficult to dynamically adapt to the requirements of different optimization stages, resulting in ignoring equipment stability in the early stage or over-pursuing process accuracy in the later stage. At the same time, their weighting methods mostly rely on subjective experience or equal distribution, being easily affected by human biases and unable to capture the short-term fluctuation characteristics of equipment operation, making it difficult to achieve coordinated optimization of process quality and equipment reliability in complex production scenarios.

[0101] The specific ways to calculate the fitness value of the position of each individual using the phased hybrid strategy method include:

[0102] For the terminal process data in the individual, calculate the absolute deviation amounts of the crimping force, insertion force, and extraction force in the terminal process data from the ideal crimping force, ideal insertion force, and ideal extraction force, and obtain the deviation degree value based on the ratio of the deviation amount to the standard range of the crimping force, insertion force, and extraction force. Among them, the ideal crimping force, ideal insertion force, and ideal extraction force are obtained through the connector industry standard.

[0103] For the assembly equipment status data in the individual, extract the fluctuation standard value of the equipment operation speed data, and calculate the fluctuation ratio of the equipment operation speed data with the reference value of the equipment operation speed data. Calculate the temperature deviation ratio of the equipment operation temperature data in the individual with the ratio of the absolute difference between the equipment operation temperature data and the reference operation temperature data to the reference operation temperature data. According to the fluctuation ratio of the equipment operation speed data and the temperature deviation ratio of the equipment operation temperature data, for example, the entropy weight method can be used to calculate the stability evaluation value. Among them, the reference operation temperature data is set by the empirical method.

[0104] When the execution times are within the first 30%, mainly use the deviation degree value and supplement with the stability evaluation value, adjust the weights of the deviation degree value and the stability evaluation value using the dynamic attenuation function, and calculate the fitness value of the individual. When the execution times are within the first 30%, the formula for using the dynamic attenuation function to adjust the weight of the deviation degree value is: hs = e -As*t , where hs is the weight of the deviation degree value, As is the attenuation coefficient, and t is the current execution times; the weight of the stability evaluation value is: ws = 1 - hs, where ws is the weight of the stability evaluation value.

[0105] When the number of executions is the latter 70%, the stability evaluation value is the main factor, and the deviation value is the auxiliary factor. The dynamic attenuation function is used to adjust the weights of the stability evaluation value and the deviation value, and the fitness value of the individual is calculated; when the number of executions is the latter 70%, the formula for using the dynamic attenuation function to adjust the weight of the stability evaluation value is: The weight of the deviation value is: hs = 1 - ws;

[0106] Among them, the fluctuation standard value of the device operation speed data is calculated using the time window method. Calculate the time window where the operation speed data is located, and calculate the maximum and minimum device operation speed data in this time window. Calculate the fluctuation standard value of the device operation speed data by taking the ratio of the device operation speed data to the maximum and minimum device operation speed data;

[0107] In the first 30% of the execution times, the key is to quickly screen out individuals that meet industry standards, ensure that the terminal process data meets the product qualification requirements, and avoid product unqualified due to excessive deviation in the terminal process data; while in the latter 70% of the execution times, it focuses on the device operation state data (stability evaluation value) because in long-term production, the design of the device operation state directly affects the persistence of the terminal process data. If the device operation is unstable (such as large speed fluctuations, abnormal temperature), even if the terminal process data meets the standards in the short term, it may still lead to quality fluctuations or downtime risks due to device abnormalities; this phased strategy balances the dual needs of short-term quality control and long-term production reliability;

[0108] This method realizes the collaborative optimization of quality and reliability through a phased hybrid strategy (focusing on process parameters in the first 30% and device state parameters in the latter 70%), combined with two-dimensional evaluation (deviation value and stability evaluation value); uses the dynamic attenuation function to objectively quantify the weights of the terminal process data and device operation state data of the same individual under different execution times; this strategy not only meets the requirements of different optimization stages, but also balances short-term process accuracy and long-term device health, providing a more scientific and adaptive fitness evaluation framework for complex production scenarios.

[0109] In the existing methods, the position update of an individual depends on the guidance of the leading individual and the influence of a random coefficient. For the handling of out-of-bounds problems, a simple bounce method or random adjustment is used. This handling method causes the updated individual position in the optimization process to exceed the set boundary range, thus unable to ensure the feasibility of the solution result in the actual problem. For example, the allowable range of the pressing force is from 100 to 500 N. If in a certain execution, the pressing force in the updated individual position becomes 600 N (exceeding the upper limit), the existing method directly sets this value to 500 N, or generates a new value between 100 and 500 with a random number for adjustment. Although this method can ensure that the value does not exceed the boundary, it does not consider the specific situation after going out of bounds, which may cause the "bounced" pressing force value to be far from the actual optimization goal and even unable to meet the actual process requirements.

[0110] The specific ways to repair the out-of-bounds characteristic variables in the updated individual position using the dual repair boundary optimization method include:

[0111] Use the np.max method and the np.min method to obtain the maximum and minimum values of the terminal process data and the assembly equipment status data in the individual position, that is, the maximum and minimum values of the characteristic variables, and use the maximum value as the upper boundary of the boundary and the minimum value as the lower boundary of the boundary.

[0112] If or Then it means that the j-th characteristic variable of the i-th updated individual position is out of bounds, where Lb j represents the upper boundary of the j-th characteristic variable, and Ub j represents the lower boundary of the j-th characteristic variable.

[0113] For the terminal process data and the assembly equipment status data in the out-of-bounds updated individual position, use the neighborhood search repair method for repair, and calculate the neighborhood radius. The formula is: where rd t,j is the neighborhood radius of the j-th characteristic variable in the t-th execution. As the execution progresses, the neighborhood radius gradually decreases. t is the current execution number, T max is the maximum execution number, and γ is the shrinkage rate parameter.

[0114] This calculation formula is used to calculate the neighborhood radius, which represents the size of the individual search range at the t-th execution. In the formula, Lb j and Ub j are the upper and lower boundaries of the characteristic variable, which determine the total size of the search space; t is the current execution number, and T maxis the maximum number of executions, and γ is the shrinkage rate parameter (taking 1 or 2); the meaning of the formula is that as the number of executions increases, the neighborhood radius gradually decreases; specifically, the size of the neighborhood radius shrinks at a certain rate with the execution process, aiming to gradually concentrate the search near the optimal solution, thereby accelerating convergence;

[0115] For the out-of-bounds terminal process data and assembly equipment status data in updating the individual position, the characteristic variables and neighborhood radius in the position of individual 1 in the current execution are used for repair to obtain the updated individual position after repair. The formula is: where α t,j represents the j-th characteristic variable in the position of individual 1 at the t-th execution, and rand(-rd t,j , rd t,j ) represents a random number randomly generated in the interval [-rd t,j , rd t,j , represents the repair value of the j-th characteristic variable in the position of the i-th updated individual; for the repair value, it is further judged whether it is out of bounds. If represents exceeding the upper limit, then the upper limit Lb j is defined as the repair value, or represents exceeding the lower limit, then the lower limit Ub j is defined as the repair value;

[0116] The neighborhood search repair method is adopted to solve the out-of-bounds problem, which has several important advantages. First, using the np.max and np.min methods to analyze the maximum and minimum values of each characteristic variable can ensure that the characteristic variables in the updated individual position always remain within the legal search range; this boundary repair method can effectively avoid the individual position exceeding the limit and enhance stability;

[0117] Then, as the execution progresses, the mechanism of gradually decreasing the neighborhood radius makes the repair process more refined and meets the actual requirements, avoiding affecting the optimization result of the search process due to excessive repair. This dynamically adjusted repair method can reduce the waste of unnecessary search space and ensure high efficiency and the quality of the final result;

[0118] Second, the neighborhood search repair method can ensure that it still retains efficient local and global search capabilities after multiple repairs during the execution process, thereby improving the solution accuracy;

[0119] In addition, by further judging and adjusting the characteristic variables repaired by the neighborhood search repair method, it can be ensured that the repaired values always remain within a reasonable boundary range, avoiding solutions that do not meet the actual requirements. This process not only improves the effectiveness and feasibility of the solution but also ensures the quality of the optimized solution, avoiding invalid or unreasonable solutions that may be caused by the simple bounce method in the existing methods;

[0120] Finally, the existing intelligent assembly optimization system for wire harnesses relies on empirical methods to set process parameters and lacks an intelligent adjustment mechanism, which in turn affects the assembly quality and production stability. By using the neighborhood search repair method and the boundary dynamic projection method, the upper and lower limits of the boundary are dynamically set using the np.max and np.min methods to ensure that the characteristic variables do not exceed the limit during the optimization process. At the same time, the out-of-bounds variables are adaptively adjusted in the later stage of optimization through the neighborhood search strategy to avoid the problem of optimization failure in the existing methods. Compared with the existing methods, this method can effectively prevent abnormal characteristic variables, improve the optimization stability, ensure that the characteristic variables fluctuate within a reasonable range, and make the assembly quality more stable and the optimization results more reliable.

[0121] The specific steps of driving the automated assembly of wire harnesses and terminals by inputting the optimal process data into the assembly equipment control system include:

[0122] Input the optimal process data, that is, the optimal terminal process data and the assembly equipment status data, into the assembly equipment control subsystem. The subsystem executes commands according to the input optimal process data to assemble and connect the wire harness and the terminal to complete the assembly.

[0123] Based on the electrical performance data and environmental disturbance data, the specific ways to construct an electrical connectivity prediction model and predict the electrical connectivity data include:

[0124] Perform missing value, outlier, and normalization processing on the electrical performance data and environmental data to generate a normalized electrical feature set;

[0125] Input M_M samples, each group of samples includes a normalized electrical feature set and the corresponding electrical connectivity data;

[0126] Use the random method to initialize the hyperparameters, weight vector, and bias term. The hyperparameters include the penalty coefficient C, epsilon parameter, and gamma parameter; set the number of executions;

[0127] Calculate the initial loss function value Among them, represents regularization, qq0 represents the initialized weight vector, pp0 represents the initialized bias term, represents the error term, q is the sample index, ξ represents the error of the qth group of samples on the upper deviation, ξ * represents the error of the qth group of samples on the lower deviation;

[0128] Use the gradient descent method to update the weight vector and bias term during each execution, and calculate the new loss function during the update process;

[0129] When the set number of executions is reached, stop the execution and output the predicted electrical connectivity data;

[0130] Predict electrical connectivity data after assembly is completed, determine whether the terminal and wire harness connections are normal, avoid subsequent manual inspection and repair, improve production efficiency and reduce error rates.

[0131] The methods for obtaining dynamic quality judgment data resistant to environmental interference include:

[0132] Based on vibration data and electromagnetic interference data, use the linear regression method to construct physical field feedback signal data, WL = 1 - a1*V - a2*EMI; where, a1 represents the influence coefficient of vibration data on electrical connectivity, that is, the vibration influence degree coefficient, V represents vibration data, a2 represents the influence coefficient of electromagnetic interference data on electrical connectivity, that is, the electromagnetic interference influence degree coefficient, EMI represents electromagnetic interference data, and WL represents physical field feedback signal data;

[0133] Based on the predicted electrical connectivity data, use the physical field feedback signal data for correction to form dynamic quality judgment data resistant to environmental interference. The formula is: DC fin = DC pre *WL, where, DC fin is the dynamic quality judgment data formed to resist environmental interference, and DC pre is the predicted electrical connectivity data;

[0134] Vibration data has a negative impact on electrical connections, especially in a frequent mechanical vibration environment. Therefore, the vibration influence degree coefficient will be a positive value. The stronger the vibration, the lower the dynamic quality judgment data resistant to environmental interference;

[0135] Electromagnetic interference data can cause signal distortion or electrical interference, especially in a high-frequency electrical system. The electromagnetic interference influence degree coefficient is also a positive value. The stronger the electromagnetic interference, the further the dynamic quality judgment data resistant to environmental interference will decrease;

[0136] By using vibration data and electromagnetic interference data to correct the predicted electrical connectivity data, the influence of external influencing factors on the electrical connectivity data can be considered more accurately; the corrected dynamic quality judgment data resistant to environmental interference is closer to the electrical behavior in actual assembly, which can effectively avoid misjudgment caused by environmental factors and improve the accuracy and reliability of the judgment result;

[0137] The specific methods for determining whether the assembly is qualified based on the dynamic quality judgment data resistant to environmental interference include:

[0138] When the dynamically generated environmental interference resistant quality judgment data is greater than 0.5, it indicates good electrical connectivity and the assembly is qualified; when the dynamically generated environmental interference resistant quality judgment data is less than or equal to 0.5, it indicates poor electrical connectivity and the assembly is unqualified; based on industry experience, using 0.5 as a median value can intuitively distinguish between good and poor.

[0139] The dynamically generated environmental interference resistant quality judgment data after correction can provide more reliable feedback under real working conditions, reduce misjudgments or missed judgments caused by external influencing factors, and improve the accuracy of product quality control; the dynamically generated environmental interference resistant quality judgment data is a continuous indicator. When its value is greater than 0.5, it usually indicates that the connection quality reaches a relatively stable and reliable level; when it is less than or equal to 0.5, it indicates poor electrical connectivity and there is a risk of unstable connection or poor contact. Therefore, it is necessary to reassemble to ensure that the electrical performance meets the standards.

[0140] The dynamically generated environmental interference resistant quality judgment data directly reflects the electrical connection quality between the terminal and the wire harness and is a key indicator for judging whether the assembly is qualified. If the connection between the terminal and the wire harness is poor, resulting in the dynamically generated environmental interference resistant quality judgment data showing a large contact resistance or poor conduction, it will affect the current conduction and the normal operation of the equipment. Therefore, by using the dynamically generated environmental interference resistant quality judgment data to judge whether the assembly is qualified, it can effectively identify whether the connection between the terminal and the wire harness is firm and reliable; if the dynamically generated environmental interference resistant quality judgment data does not meet the standard, it indicates that there is a problem with the connection and it is necessary to reassemble to ensure the assembly quality and the subsequent normal operation of the equipment; thus avoiding the problem that the existing intelligent assembly optimization system can only detect whether the assembly is qualified in the subsequent detection link.

[0141] In this embodiment, the multi-source data sensing subsystem collects and monitors terminal process data, assembly equipment status data, electrical performance data, environmental data, vibration data, and electromagnetic interference data in real time, and the system can perform real-time perception and feedback on each link in the assembly process;

[0142] The process parameter optimization engine is based on a population search optimization mechanism. Through a phased hybrid strategy method and a dual repair boundary optimization method, it precisely optimizes each process parameter, such as the pressing force, insertion force, and pulling force, to ensure the best efficiency and stability of the assembly process, thereby significantly improving the assembly accuracy and reducing assembly problems caused by improper process settings;

[0143] The precision assembly execution controller precisely controls the assembly process of the wire harness and the terminal according to the optimal process data to ensure high precision and high stability of the assembly;

[0144] Electrical connectivity prediction is carried out by combining electrical performance data and environmental data, and the prediction results are further corrected by using vibration data and electromagnetic interference data, so as to accurately judge whether the assembly quality is qualified. This process can not only predict the assembly results in advance, but also adjust according to external factors, avoiding the neglect of external interference factors by the detection method and improving the reliability and long-term performance of the assembly.

[0145] Embodiment 2

[0146] Please refer to Figure 3 As shown, for the parts not described in detail in this embodiment, refer to the description in Embodiment 1. A method for optimizing intelligent assembly of wire harnesses is provided, including:

[0147] Step SS1: Real-time capture terminal process data, assembly equipment status data, electrical performance data, environmental data, vibration data, and electromagnetic interference data;

[0148] Step SS2: Relying on the terminal process data and assembly equipment status data, generate the optimal process data through a population search optimization mechanism; in the population search optimization mechanism, use a phased hybrid strategy to calculate the fitness value, update the position of the leading individual and correct the out-of-bounds variables by combining the double repair boundary optimization method, and execute to output the optimal process data;

[0149] Among them, the phased hybrid strategy for calculating the fitness value is to calculate the deviation value and stability evaluation value in the individual; according to the number of executions, calculate the fitness value of each individual by using the deviation value and stability evaluation value;

[0150] Among them, in the double repair boundary optimization method, the neighborhood search repair method is used to repair the updated individual position; then the boundary dynamic projection method is used to repair the updated individual position after repair;

[0151] Step SS3: Input the optimal process data into the assembly equipment control system to drive the automatic assembly of the wire harness and terminals;

[0152] Step SS4: For the assembled wire harness, based on the electrical performance data and environmental disturbance data, construct an electrical connectivity prediction model to predict the electrical connectivity data; fuse the vibration data and electromagnetic interference data to construct the physical field feedback signal data, and correct the predicted electrical connectivity data to form the dynamic quality judgment data resistant to environmental interference;

[0153] Step SS5: Judge whether the assembly is qualified based on the dynamic quality judgment data resistant to environmental interference; if the dynamic quality judgment data resistant to environmental interference is good, the assembly is qualified; if the dynamic quality judgment data resistant to environmental interference is poor, the assembly is unqualified, dynamically adjust the double repair boundary optimization method in the population search mechanism to regenerate the new optimal process data, and perform the assembly according to the new optimal process data.

[0154] Example 3

[0155] This embodiment discloses an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it implements the operation mode of a wire harness intelligent assembly optimization platform and method provided above.

[0156] Since the electronic device introduced in this embodiment is the electronic device adopted for implementing a wire harness intelligent assembly optimization platform and method in the embodiments of the present application, based on the wire harness intelligent assembly optimization platform and method introduced in the embodiments of the present application, those skilled in the art can understand the specific implementation manners and various variations of the electronic device in this embodiment. Therefore, the detailed introduction of how the electronic device implements the method in the embodiments of the present application will not be repeated here. As long as those skilled in the art implement the electronic device adopted for a wire harness intelligent assembly optimization platform and method in the embodiments of the present application, it falls within the scope of protection of the present application.

[0157] The above formulas are all calculated by taking the numerical values after dimensionless. The formula is a formula obtained by collecting a large amount of data for software simulation to approximate the real situation. The preset parameters and threshold selection in the formula are set by those skilled in the art according to the actual situation.

[0158] The above are only the preferred embodiments of the present invention. The protection scope of the present invention is not limited to the above embodiments. All technical solutions within the idea of the present invention belong to the protection scope of the present invention. It should be noted that for ordinary technical users in the technical field, several improvements and refinements made without departing from the principle of the present invention should also be regarded as within the protection scope of the present invention.

Claims

1. An intelligent assembly optimization platform for wire harnesses, characterized in that, Including: Multi-source data sensing subsystem: capturing terminal process data, assembly equipment status data, electrical performance data, environmental data, vibration data, and electromagnetic interference data in real time; Process parameter optimization engine subsystem: relying on terminal process data and assembly equipment status data, generating optimal process data through a population search optimization mechanism; In the population search optimization mechanism, a phased hybrid strategy is used to calculate the fitness value, and the out-of-bounds variables are corrected by combining the leading individual position update with the dual repair boundary optimization method, and the optimal process data is output; Among them, calculating the fitness value by the phased hybrid strategy is to calculate the deviation value and stability evaluation value in the individual; according to the number of executions, the fitness value of each individual is calculated using the deviation value and stability evaluation value; Among them, in the dual repair boundary optimization method, the neighborhood search repair method is used to repair the updated individual position; then the boundary dynamic projection method is used to repair the updated individual position after repair; Precision assembly execution control subsystem: inputting the optimal process data into the assembly equipment control system to drive the automated assembly of the wire harness and terminals; Anti-environmental interference dynamic quality evaluation subsystem: for the assembled wire harness, based on electrical performance data and environmental disturbance data, constructing an electrical connectivity prediction model to predict electrical connectivity data; fusing vibration data and electromagnetic interference data to construct physical field feedback signal data, and correcting the predicted electrical connectivity data to form dynamic quality judgment data against environmental interference; Real-time monitoring subsystem: judging whether the assembly is qualified based on the dynamic quality judgment data against environmental interference; if the dynamic quality judgment data against environmental interference is good, the assembly is qualified; if the dynamic quality judgment data against environmental interference is poor, the assembly is unqualified, and the dual repair boundary optimization method in the population search mechanism is dynamically adjusted to regenerate new optimal process data, and the assembly is carried out according to the new optimal process data.

2. The intelligent assembly optimization platform for wire harnesses according to claim 1, characterized in that, The terminal process data includes crimping force data, insertion force data, and extraction force data; The assembly equipment status data includes equipment running speed data and equipment running temperature data; The electrical performance data includes voltage data, current data, contact resistance data, conduction resistance data, and insulation resistance data; The environmental data includes environmental temperature data and environmental humidity data.

3. The intelligent assembly optimization platform for wire harnesses according to claim 2, characterized in that The specific method of generating optimal process data through the population search optimization mechanism relying on terminal process data and assembly equipment status data includes: Step A1: Regarding a set of terminal process data and assembly equipment status data as an individual; setting the number of individuals; setting the maximum number of executions; Step A2: Setting a standard range for the terminal process data and the status data of the transfer equipment, and randomly initializing the data values within the standard range as the positions of the individuals; Step A3: Using the phased hybrid strategy method to calculate the fitness value of the position of each individual; Step A4: During the execution process, select the leading individuals and perform position updates; the leading individuals are the individuals with the top 3 fitness values, including individuals 1, 2, and 3; Step A5: Update according to the positions of the leading individuals to obtain the updated individual positions; Step A6: After the position update is completed, use the double repair boundary optimization method to repair the out-of-bounds characteristic variables in the updated individual position; Step A7: Repeat Steps A4 to A6 until the set maximum number of executions is reached, and output a set of optimal process data, namely the optimal terminal process data and assembly equipment status data.

4. The intelligent assembly optimization platform for wire harnesses according to claim 3, characterized in that The specific method of using the phased hybrid strategy method to calculate the fitness value of the position of each individual includes: For the terminal process data in the individual, calculate the absolute deviation amounts of the pressing force, insertion force, and pulling force in the terminal process data from the ideal pressing force, ideal insertion force, and ideal pulling force, and obtain the deviation degree value based on the ratio of the deviation amount to the standard ranges of the pressing force, insertion force, and pulling force; For the assembly equipment status data in the individual, extract the fluctuation standard value of the equipment operating speed data, and calculate the fluctuation ratio of the equipment operating speed data with the reference value of the equipment operating speed data; calculate the temperature deviation ratio of the equipment operating temperature data in the individual by taking the absolute difference between the equipment operating temperature data and the reference operating temperature data and dividing it by the reference operating temperature data; calculate the stability evaluation value based on the fluctuation ratio of the equipment operating speed data and the temperature deviation ratio of the equipment operating temperature data; When the number of executions is the first 30%, mainly based on the deviation degree value and supplemented by the stability evaluation value, use the dynamic attenuation function to adjust the weights of the deviation degree value and the stability evaluation value, and calculate the fitness value of the individual; When the number of executions is the last 70%, mainly based on the stability evaluation value and supplemented by the deviation degree value, use the dynamic attenuation function to adjust the weights of the stability evaluation value and the deviation degree value, and calculate the fitness value of the individual.

5. The intelligent assembly optimization platform for wire harnesses according to claim 4, characterized in that, The specific method of using the double repair boundary optimization method to repair the out-of-bounds characteristic variables in the updated individual position includes: Use the np.max method and np.min method to obtain the maximum and minimum values of the terminal process data and assembly equipment status data in the individual position, that is, the maximum and minimum values of the characteristic variables, and use the maximum value as the upper boundary and the minimum value as the lower boundary; If the j-th characteristic variable of the i-th updated individual position is greater than the upper boundary of the j-th characteristic variable, or the j-th characteristic variable of the i-th updated individual position is less than the lower boundary of the j-th characteristic variable, it means that the j-th characteristic variable of the i-th updated individual position is out of bounds; For the out-of-bounds terminal process data and assembly equipment status data in the updated individual position, use the neighborhood search repair method for repair. First, calculate the neighborhood radius based on the difference between the upper and lower limits, the ratio of the current execution times to the maximum execution times, and the shrinkage rate parameter; For the out-of-bounds terminal process data and assembly equipment status data in the updated individual position, use the terminal process data and assembly equipment status data in the position of individual 1 in the current execution and the neighborhood radius to perform repair to obtain the repaired updated individual position; for the repaired value in the updated individual position, use the boundary dynamic projection method to further judge whether it is out of bounds. If it exceeds the upper limit, define the upper limit as the repaired value. If it exceeds the lower limit, define the lower limit as the repaired value.

6. The intelligent assembly optimization platform for wire harnesses according to claim 5, characterized in that The specific steps of inputting the optimal process data into the control system of the assembly equipment to drive the automated assembly of the wire harness and terminals are as follows: Input the optimal process data, namely the optimal terminal process data and the assembly equipment status data, into the control subsystem of the assembly equipment. The subsystem executes commands according to the input optimal process data to assemble and connect the wire harness and terminals.

7. The intelligent assembly optimization platform for wire harnesses according to claim 6, characterized in that The specific method of constructing an electrical connectivity prediction model based on electrical performance data and environmental disturbance data to predict electrical connectivity data includes: Perform missing value, outlier, and normalization processing on the electrical performance data and environmental data to generate a normalized electrical feature set; Input M_M samples, each group of samples including a normalized electrical feature set and the corresponding electrical connectivity data; Use the random method to initialize the hyperparameters, weight vectors, and bias terms. The hyperparameters include the penalty coefficient C, epsilon parameter, and gamma parameter; set the number of executions; Calculate the initial loss function value of the model; Use the gradient descent method to update the weight vectors and bias terms during each execution, and calculate the new loss function during the update process; When the set number of executions is reached, stop the execution and output the predicted electrical connectivity data.

8. The intelligent assembly optimization platform for wire harnesses according to claim 7, characterized in that The acquisition method of the dynamic quality judgment data for anti-environmental interference includes: Use vibration data and electromagnetic interference data to construct physical field feedback signal data; based on the predicted electrical connectivity data, use the physical field feedback signal data for correction to obtain the dynamic quality judgment data for anti-environmental interference.

9. The intelligent assembly optimization platform for wire harnesses according to claim 8, characterized in that, The specific method of judging whether the assembly is qualified based on the dynamic quality judgment data for anti-environmental interference includes: When the dynamic quality judgment data for anti-environmental interference is greater than 0.5, it indicates good electrical connectivity, and the assembly is qualified; when the dynamic quality judgment data for anti-environmental interference is less than or equal to 0.5, it indicates poor electrical connectivity, and the assembly is unqualified.

10. A method for optimizing the intelligent assembly of a wire harness, which is applied to the wire harness intelligent assembly optimization platform according to any one of claims 1 to 9, and is characterized in that, Including: Step SS1: Real-time capture of terminal process data, assembly equipment status data, electrical performance data, environmental data, vibration data, and electromagnetic interference data; Step SS2: Rely on the terminal process data and assembly equipment status data to generate optimal process data through a population search optimization mechanism; in the population search optimization mechanism, use a phased hybrid strategy to calculate the fitness value, update the position of the leading individual and correct the out-of-bounds variables using the double repair boundary optimization method, and execute to output the optimal process data; Among them, the phased hybrid strategy for calculating the fitness value is to calculate the deviation value and stability evaluation value in the individual; according to the number of executions, calculate the fitness value of each individual using the deviation value and stability evaluation value; Among them, in the double repair boundary optimization method, the neighborhood search repair method is used to repair the updated individual position; then the boundary dynamic projection method is used to repair the updated individual position after repair; Step SS3: Input the optimal process data into the control system of the assembly equipment to drive the automated assembly of the wire harness and terminals; Step SS4: For the assembled wire harness, based on the electrical performance data and environmental disturbance data, construct an electrical connectivity prediction model to predict the electrical connectivity data; fuse the vibration data and electromagnetic interference data to construct the physical field feedback signal data, and correct the predicted electrical connectivity data to form the dynamic quality judgment data resistant to environmental interference. Step SS5: Judge whether the assembly is qualified based on the dynamic quality judgment data resistant to environmental interference; if the dynamic quality judgment data resistant to environmental interference is good, the assembly is qualified; if the dynamic quality judgment data resistant to environmental interference is poor, the assembly is unqualified, and the dual repair boundary optimization method in the population search mechanism is dynamically adjusted to regenerate new optimal process data, and the assembly is carried out according to the new optimal process data.

Images