Intelligent winding and detecting method and system for three-phase integrated coil of soft start device

Abstract

The invention provides an intelligent winding and detecting method and system for a three-phase integrated coil of a soft start device, and relates to the technical field of winding of the three-phase integrated coil, and the method comprises the steps: determining the electrical parameters of coil winding, and preparing a winding material; calculating the optimal tension within the T moment by establishing a tension mathematical model to adaptively control the tension of the wire; performing real-time correction and compensation on the key parameters by using a neural network model; the visual detection control center detects the coil in real time and judges whether errors occur in the winding process, and if yes, winding is stopped, and an alarm is given out; and putting the wound coil into epoxy resin glue for defoaming to ensure that the insulating material covers a gap between the surface of the wire and the coil. The wire tension is adaptively adjusted by using the model prediction control algorithm and the neural network model, so that the wire tension is controlled within the preset range, the wire is prevented from being damaged due to excessive stretching or loosening, the integrity and stability of the wire are favorably kept, and the overall performance of the soft start device is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of three-phase integration, and in particular to a three-phase integration coil intelligent winding and detection method for a soft starter device, and a three-phase integration coil intelligent winding and detection system for a soft starter device. Background Art

[0002] The three-phase integrated coil winding of the soft starter device refers to winding the wire into a three-phase integrated coil according to specific rules and parameters, and applying it to the soft starter. This winding method enables the coil to produce smooth magnetic field changes when energized, thereby achieving smooth starting and speed control of the motor.

[0003] If the wire tension is not properly controlled, it may have a serious impact on the performance and service life of the soft starter. Too much tension may cause problems such as coil deformation and insulation damage; while too little tension may cause the coil to loosen and electrical performance to deteriorate. These problems will directly affect the normal operation and service life of the soft starter.

[0004] When the wire tension is too large, the coil may be subjected to excessive mechanical stress during the winding process, causing the coil shape to change. This deformation may affect the mechanical strength and electrical performance of the coil, and even cause the coil to fail to work properly. It may also cause the wire insulation layer to be damaged. During the winding process, if the wire is subjected to excessive tension, the insulation layer may be damaged due to friction or stretching, which will reduce the electrical insulation performance of the coil and increase the risk of electrical failure. If the wire tension is too small, the coil may not be able to remain tightly arranged during the winding process, resulting in a loose coil. This loose state may affect the mechanical strength and heat dissipation performance of the coil and reduce the service life of the soft starter. Summary of the invention

[0005] The invention provides a three-phase integrated coil intelligent winding and detection method for a soft starter device, which is used to solve the defect in the prior art that the wire tension is improperly controlled during the winding process, thus affecting the performance and life of the soft starter device.

[0006] The present invention provides a three-phase integrated coil intelligent winding and detection method for a soft start device, comprising: S1: Determine the electrical parameters of the coil winding and prepare the winding materials; S2: During the coil winding process, the conductor tension is adaptively controlled, and the optimal tension within time T is calculated by establishing a tension mathematical model; S3: Use the neural network model to estimate and correct the key parameters in the tension mathematical model in real time and compensate for the nonlinear part; S4: The visual inspection control center detects the coil winding quality in real time and determines whether there is an error in the winding process. If so, the winding process is stopped and an alarm is issued; S5: Place the wound coil into epoxy resin glue for degassing, ensuring that the insulating material is evenly covered on the surface of the wire and in the gap between the coil.

[0007] According to a three-phase integrated coil intelligent winding and detection method for a soft starter device provided by the present invention, in step S2, the specific steps of adaptively controlling the wire tension are: S21: Using a tension sensor to collect the tension of the wire to obtain tension data; S22: preprocessing the tension data to obtain tension preprocessing data; S23: Use the model predictive control algorithm to establish a tension mathematical model for the coil winding system and calculate the optimal tension within time T.

[0008] According to a soft starter device three-phase integrated coil intelligent winding and detection method provided by the present invention, in step S22, preprocessing includes data normalization, wild point elimination, and periodic line inspection; Normalize the data and scale it to a specific range; The IQR method is used to calculate wild points. The degree of data dispersion is determined by calculating the quartiles of the tension training data, and whether it is a wild point is determined based on the set threshold. The steps for calculating wild points using the IQR method are: Sort the data set by numerical size; Calculate the first quartile Q1. In the sorted data set, Q1 is the data point at the 25% position. Calculate the third quartile Q3. In the sorted data set, Q3 is the data point at the 75% position. Calculate IQR, IQR=Q3-Q1; After obtaining the IQR, the set threshold is used to determine whether the data point is an outlier. If so, the data point is removed. Perform periodic checks on the data to ensure that there is no duplicate data in the data set.

[0009] According to a three-phase integrated coil intelligent winding and detection method for a soft starter device provided by the present invention, in step S23, a tension mathematical model is established for the coil winding system using a model predictive control algorithm, and the specific steps for calculating the optimal tension within time T are: S231: In the coil winding system, determine the input variables and output variables. The input variables are the motor speed, motor torque, wire diameter and number of turns, and the output variable is the wire tension; S232: Establishing a tension mathematical model based on physical principles to obtain a tension change prediction formula; S233: predicting future tension changes according to a tension change prediction formula; S234: Set the input constraint conditions, output constraint conditions and objective function of the wire tension during the test process. Use the quadratic programming algorithm to calculate the optimal output torque and speed of the motor within time T.

[0010] According to a three-phase integrated coil intelligent winding and detection method for a soft starter device provided by the present invention, in step S232, the tension change prediction formula is expressed as: Where F is the wire tension, T is the torque, T0 is the other resistance torque, d is the wire diameter, and N is the number of turns.

[0011] According to a three-phase integrated coil intelligent winding and detection method for a soft starter device provided by the present invention, in step S234, the optimal torque obtained by solving the quadratic programming is expressed as: In the formula, is the optimal control sequence at time k.

[0012] According to a three-phase integrated coil intelligent winding and detection method for a soft starter device provided by the present invention, in step S3, the neural network model compensates for the nonlinear part that is difficult to accurately model by the tension mathematical model predictive control algorithm, including: There is an error between the conductor tension prediction value calculated by the prediction model algorithm and the conductor tension actually measured.

[0013] The neural network model and the model predictive control algorithm work in parallel, and the same input is input into the tension change model formula of the model predictive control algorithm and the neural network model respectively. The model predictive control algorithm outputs the prediction result according to its own linear relationship, and the neural network model outputs the corresponding nonlinear compensation amount according to the nonlinear mapping relationship it has learned. The linear relationship prediction result of the model predictive control algorithm is fused with the nonlinear relationship prediction result of the neural network to obtain a final prediction result of linear and nonlinear factors.

[0014] According to a method for intelligent winding and detection of a three-phase integrated coil of a soft starter device provided by the present invention, in step S3, the specific steps of outputting a nonlinear compensation amount of a nonlinear mapping relationship of a neural network model include: S31: Select a multilayer perceptron to predict the nonlinear compensation amount of tension, and initialize the connection weights and bias items between neurons in each layer of the neural network to obtain initialization data; S32: The neurons in the hidden layer perform weighted summation on the initialization data and perform nonlinear transformation through an activation function to obtain a predicted value of a nonlinear compensation amount of the initialization data; S33: using the mean square error to calculate the loss function, and reducing the loss of the predicted value of the next nonlinear compensation amount; S34: Use the chain rule to calculate the gradients of the connection weights and biases of the neurons in each layer in reverse order from the output layer, and use the stochastic gradient descent optimization algorithm to update the parameters in the neural network based on the gradients; S35: Repeat the weighted summation, loss function calculation and parameter update steps until the neural network model can stably output a reasonable and accurate nonlinear compensation amount.

[0015] According to a soft starter three-phase integrated coil intelligent winding and detection method provided by the present invention, in step S4, the specific steps of using a visual detection control center to monitor the winding quality in real time are: S41: The visual inspection control center collects coil images during the winding process, processes the coil images using an image analysis algorithm, and extracts image features; S42: judging that the winding quality meets the requirements according to the image features and the preset quality judgment parameters, and if so, continuing to collect and analyze the next set of images; S43: After receiving the winding quality problem information fed back by the visual inspection control center, the winding machine control center will immediately trigger a stop command to stop the winding machine; S44: After the winding machine stops running, the winding machine control center issues an alarm.

[0016] The present invention also provides a three-phase integrated coil intelligent winding and detection system for a soft starter device, comprising: an electrical parameter module for determining electrical parameters of coil winding; The adaptive winding module is used to adaptively control the wire tension during the coil winding process and calculate the optimal tension within the T time by establishing a tension mathematical model; A tension data acquisition unit, used to acquire the tension of the conductor using a tension sensor to obtain tension data; A tension data preprocessing unit, used for preprocessing tension data; A tension calculation unit, used to calculate the optimal tension within time T using a model predictive control algorithm; Nonlinear compensation module, used to use neural network model to estimate and correct key parameters in tension mathematical model in real time and compensate for nonlinear part; The winding quality detection module is used to detect the coil winding quality in real time using the visual detection control center to determine whether there is an error in the winding process. If so, the winding process is stopped and an alarm is issued; The post-winding processing module is used to place the wound coil into epoxy resin glue for degassing and to protect the wound coil.

[0017] The three-phase integrated coil intelligent winding and detection method for a soft starter provided by the present invention establishes a tension mathematical model for the coil winding system by using a model predictive control algorithm, calculates the optimal tension in a limited time domain, and uses a neural network model to estimate and correct key parameters in the model prediction in real time, and compensates for the nonlinear part that is difficult to accurately model by the model predictive control algorithm, thereby solving the defect that improper wire tension control affects the performance and life of the soft starter during the winding process, and the beneficial effects achieved are: MPC is good at handling constraint problems and model-based dynamic prediction, while the neural network model can deal with nonlinear and complex unknown relationships. The combination of the two gives full play to their respective advantages, improves the adaptability of the entire tension control system to complex working conditions, and can control tension more accurately. Through the assistance of the neural network model to MPC, the prediction model can be more in line with the actual coil winding system, reducing the tension control deviation caused by model errors and incompletely considered nonlinear factors, thereby ensuring the stability and uniformity of tension throughout the winding process and avoiding coil winding quality problems caused by improper tension.

[0018] The neural network model can estimate and correct the key parameters in the model predictive control algorithm in real time, thereby ensuring the accuracy and reliability of the tension mathematical model. The neural network model more accurately predicts the optimal tension in the future limited time domain, reducing the tension control error caused by inaccurate parameters. The tension control in the coil winding system is often affected by a variety of nonlinear factors, such as material properties, friction resistance, etc. The neural network model can compensate for these nonlinear factors, thereby improving the accuracy and stability of tension control. Combining the neural network model with the model predictive control algorithm can achieve a higher level of automated control, reduce manual intervention, and improve production efficiency and product quality. The powerful learning and adaptability of the neural network model helps to enhance the adaptability and flexibility of the system to meet the needs of different application scenarios.

[0019] Precise tension control helps reduce tension fluctuations during the winding process, thereby avoiding damage to the wire due to excessive stretching or relaxation, maintaining the integrity and stability of the wire, and improving the overall performance of the soft starter. By precisely controlling the tension, the friction and wear between the wire and the winding mechanism are reduced, thereby reducing the risk of fatigue damage to the wire, extending the service life of the wire, and reducing the frequency and cost of replacing the wire.

[0020] Precise tension control helps ensure that the wire maintains uniform tension distribution during the winding process, thereby improving the quality and consistency of the winding, enhancing the performance and reliability of the soft starter, and reducing the failure rate caused by winding quality problems. By accurately controlling the tension, the downtime caused by improper tension control is reduced, thereby improving the operating efficiency of the overall system.

[0021] By precisely controlling the winding speed and tension, the insulation and electrical performance of the coil are guaranteed. Optimizing the layout, through optimization calculation, the best layout of the winding is achieved, electromagnetic coupling and loss are reduced, the efficiency of the equipment is improved, and production efficiency is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0023] Figure 1 It is a flow chart of a three-phase integrated coil intelligent winding and detection method for a soft start device provided by an embodiment of the present invention; Figure 2 It is a flow chart of a three-phase integrated coil intelligent winding and detection method for a soft start device provided by an embodiment of the present invention; Figure 3 It is a module of a three-phase integrated coil intelligent winding and detection system for a soft starter provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0025] Combine the following Figure 1-Figure 3 The invention describes a three-phase integrated coil intelligent winding and detection method and system for a soft start device.

[0026] Figure 1 The invention discloses a three-phase integrated coil intelligent winding and detection method for a soft start device provided by an embodiment of the present invention.

[0027] like Figure 1-Figure 2 As shown, the three-phase integrated coil intelligent winding and detection method of the starting device provided by the embodiment of the present invention mainly includes the following steps: The specific steps of the intelligent winding method include: S1: Use professional electromagnetic design software to accurately design the three-phase integrated coil and determine the coil's number of turns, wire diameter, winding arrangement, connection method and other parameters.

[0028] Prepare the insulation paper, insulation sleeve, high-strength insulation material and binding wire before winding; Polyester film insulation paper is used to provide insulation between coil layers and between the coil and the inner tube to prevent short circuit. The thickness is determined according to the voltage level and insulation requirements. Prepare insulating sleeves to wrap the lead wires to avoid insulation damage during connection or subsequent operations. Use epoxy resin glue to encapsulate and fix the coils to ensure that they have good insulation, temperature resistance and adhesion; The inner cylinder is made of glass fiber reinforced epoxy resin material. The size of the inner cylinder should be precisely processed according to the size requirements of the coil, and its diameter and length should ensure that the coil has appropriate space and electrical performance after winding; The binding wire is used to temporarily fix the conductor during the winding process; the sandpaper is used to polish the surface of the inner tube to ensure its smoothness and avoid scratching the insulation layer of the conductor; Use copper wire to wind the coil to ensure the wire has good conductivity, mechanical strength and insulation performance. Debug and calibrate the intelligent winding machine to check whether the various components of the winding machine are operating normally, such as the pay-off device, the wire arrangement device, the tension control system, the counter, etc. Ensure that the accuracy and stability of the winding machine meet the winding requirements.

[0029] S2: During the coil winding process, the wire tension is adaptively controlled, and the optimal tension within time T is calculated by establishing a tension mathematical model.

[0030] Install the wire reel firmly on the pay-off stand, adjust the height and position of the pay-off stand so that the wire is at the same level as the main shaft of the winding machine to facilitate the lead-out of the wire. Install the wire clamp between the winding machine and the pay-off stand, clamp the electromagnetic wire, and adjust the clamping force of the wire clamp so that the wire has appropriate tension when winding. Leave an appropriate length at the beginning of the wire, generally 20-30 cm, depending on the specific wiring requirements and installation location. Fix the beginning end on the main shaft of the winding machine in the prescribed manner, and use winding to ensure that the beginning end will not loosen during the winding process.

[0031] S21: using a tension sensor to collect tension data of the wire to obtain tension training data; S22: preprocessing the tension training data to obtain tension preprocessing data, the preprocessing including removing wild points at data normalization, periodicity test and fixed change trend test;

[0032] Data normalization is to scale the data to a specific range in order to eliminate the differences between different data scales so that the subsequent neural network model learning can better process the data. Use Z-Score standardization to standardize the data so that the processed tension training data conforms to the standard normal distribution. The formula is expressed as: Where μ is the mean, σ is the standard deviation, and X is the tension training data.

[0033] The purpose of removing outliers is to reduce noise and outliers in the data set and improve the accuracy and reliability of the data. The IQR method is used to identify outliers. The degree of data dispersion is determined by calculating the quartiles of the tension training data, and whether it is an outlier is determined based on the set threshold.

[0034] The calculation steps of IQR are: Sort the data set by numerical size; Calculate the first quartile Q1, the minimum value of 25% of the data points in the data set. In the sorted data set, Q1 is the data point at the 25% position; Calculate the third quartile Q3, the minimum value of 75% of the data points in the data set. In the sorted data set, Q3 is the data point at the 75% position; Calculate IQR, which is the difference between Q3 and Q1, IQR=Q3-Q1; After obtaining the IQR, the data point is judged as an outlier based on the set threshold. The data point below Q1-1.5×IQR or above Q3+1.5×IQR is an outlier.

[0035] The purpose of the periodicity test is to determine whether there is a repetitive pattern or periodic change in the data set. The autocorrelation function is used to test the periodicity of the data: the periodicity of the data is determined by calculating the correlation between the data point and its own different lag periods. If the autocorrelation function has a significant peak at a certain lag period, it may indicate that the data is periodic.

[0036] The purpose of the fixed change trend test is to determine whether there is a fixed change trend in the data set. Use trend line fitting to test the data for a fixed change trend: use linear regression, polynomial regression and other methods to fit the trend line of the data, and observe the shape and slope of the trend line to determine the trend of the data.

[0037] S23: Use the model predictive control algorithm to establish a tension mathematical model for the coil winding system and calculate the optimal tension within time T.

[0038] Determine the input variables and output variables in the coil winding system.

[0039] The input variables include: motor speed n (affecting winding speed), motor torque T (directly related to tension adjustment), wire diameter and number of turns. These variables are adjusted by the controller, which in turn affects the tension during coil winding.

[0040] The output variables include: the actual tension value F. The tension is directly related to the quality of coil winding. If the tension is too large, the wire may be damaged, and if the tension is too small, the coil will not be wound tightly.

[0041] The model is established based on physical principles, and the mathematical model of tension is derived according to the principles of mechanics and kinematics.

[0042] Consider the relationship between the rotation and tension of the coil: Based on the relationship between tension and centripetal force and the relationship between linear velocity and angular velocity, it is deduced that , combining the above relationship with factors such as motor torque, an equation between the variables is established.

[0043] The coil radius r will increase with the increase of the number of winding turns N, and its rate of change is related to the winding speed, the width b of the coil, etc. The rate of change of the radius r r' is expressed as: , where b is the width of the coil.

[0044] After sorting and discretization, the discrete time state space model is obtained as follows:

[0045] in is the state vector, is the input vector, is the output vector, and A, B, C, and D are the corresponding coefficient matrices.

[0046] Based on the prediction model, the tension change in the future limited time domain is predicted in each control cycle. Assuming that during the coil winding process, the coil diameter is detected to be a certain value and the motor speed is in a specific state at the current moment, these data are input into the prediction model. The model can then calculate how the tension will change in the future as the winding continues based on its internal mathematical relationships and parameters.

[0047] Assuming the dependent variable is the wire tension, and the independent variables are the motor speed, torque, wire diameter and number of turns, the tension change prediction formula is obtained, which is expressed as: Where F is the wire tension, T is the torque, is the other resistance torque, d is the wire diameter, and N is the number of turns.

[0048] There are various constraints in the actual system, such as the upper and lower limits of the torque that the motor can provide, which limits the range of tension adjustment; the unwinding and rewinding speeds cannot exceed the safe operating speed range of the equipment; and the growth of the roll diameter is also within a certain physically feasible range.

[0049] Enter the constraints: In the formula, i=0,1,…,Nc.

[0050] Output constraints: In the formula, i=0,1,…,Np.

[0051] Clearly define the goal of tension control and keep the tension stable within a set optimal range. For example, for a certain high-precision electronic coil winding, the tension error is required to be controlled within ±1N. At the same time, it may also be expected that the rate of tension change should also be maintained within a certain reasonable range to avoid sudden changes in tension affecting the winding quality.

[0052] The objective function is expressed as: Where q is the output weight, r is the control increment weight, y(k) is the tension, r(k) is the tension reference value, u(k) is the control motor torque, N p is the predicted time domain length, N c To control the time domain length, and N c ≤N p , Represents the control increment, that is, the change in motor torque at each moment relative to the previous moment.

[0053] Combined with the set objective function and constraints, the quadratic programming algorithm is used to calculate the optimal control sequence from the current moment to the future finite time domain. During the three-phase integrated coil winding process, based on the predicted tension change, tension objective function and constraints, the optimization algorithm is used to calculate the optimal sequence of control quantities such as torque and speed that the motor should output at each moment in the next few seconds, so that the tension can be as close as possible and maintained within the preset optimal range.

[0054] The formula for calculating the optimal torque through quadratic programming is expressed as: Where U is the optimal control sequence at time k.

[0055] The first element is directly extracted from the optimal control sequence vector obtained by solving, which is the optimal torque value at the current moment k. This value is the result of minimizing the objective function given by the optimization algorithm after comprehensively considering the current state of the system, the expected tension and various constraints.

[0056] For the optimal torque of k+i, we can also get the optimal control sequence from Extract the corresponding elements, the expression is: Pick The element with index i in the vector corresponds to the optimal torque value at time k+i. It reflects the optimal torque that should be adopted at each future moment, which is pre-calculated by the optimization algorithm based on the information at the current moment k. This is used to control the motor so that the tension of the system can be as close as possible to and maintained within the set optimal range during future operation, while satisfying various torque-related constraints.

[0057] S3: The neural network model estimates and corrects the key parameters in the tension mathematical model in real time and compensates for the nonlinear part.

[0058] Compare the actual tension value with the predicted tension value output by the prediction model, and calculate the deviation between the two. If the deviation exceeds the allowable range, it means that the prediction model may not match the actual system, which may be due to unconsidered interference factors during the operation of the equipment, or the model itself has certain modeling errors. Use this deviation information to correct the prediction model so that the subsequent prediction results can be closer to the actual tension changes, thereby improving the accuracy of control.

[0059] The neural network model has a powerful nonlinear mapping capability. It uses the observable input and output data of the system as training samples, mines the implicit relationships therein, and then estimates these key parameters. The neural network model performs complex nonlinear operations in the hidden layer through its internal neuron connections and activation functions, and finally outputs the estimated values ​​related to the key parameters in the output layer, and continuously corrects them according to the actual situation to ensure the accuracy of the estimation. In each control cycle of MPC, there will be a deviation between the predicted system output calculated by the model prediction algorithm and the actual measured system output. This deviation information is fed back to the neural network model to further adjust the parameters such as the weight and threshold of the neural network model, thereby realizing the correction of the estimated values ​​of the key parameters.

[0060] The neural network model works in parallel with the linear prediction model of MPC, and the same input is sent to the linear prediction model of MPC and the neural network model respectively. The linear prediction model of MPC outputs the prediction result according to its own linear relationship, while the neural network model outputs the corresponding nonlinear compensation amount according to the nonlinear mapping relationship it has learned. The two results are properly fused to obtain a final prediction result that comprehensively considers linear and nonlinear factors for subsequent optimization and control decisions.

[0061] S31: Select a multilayer perceptron as the nonlinear compensation for predicting tension based on the current input. The multilayer perceptron includes an input layer, several hidden layers, and an output layer. For example, the number of nodes in the input layer includes five features such as speed and material properties. The input layer has five nodes, the hidden layer has 2-3 layers, and the number of neurons in each layer is set to 10 and 8 neurons respectively, and the output layer has one node. Initialize the connection weights and bias items between the neurons in each layer of the neural network. These initial parameters determine the initial computing power and expression power of the network, which will be gradually optimized through training in the future.

[0062] S32: Send the initialization data to the input layer of the neural network. The data is passed from the input layer to the hidden layer in sequence. In each hidden layer, the neurons will perform weighted summation on the initialization data and perform nonlinear transformation through the activation function. For the ReLU activation function, for a neuron j in the hidden layer, the formula for the weighted summation of the input x is expressed as: In the formula, ω ij is the weight connecting the previous layer node i to the neuron j, x i is the input value of the previous layer, b j is the bias of this neuron.

[0063] Then through the activation function Get the output a j , this output will be used as the input of the next layer of neurons, and the calculation will be continued layer by layer to achieve feature extraction and nonlinear transformation of the input data.

[0064] After multiple calculations in the hidden layer, the data is finally passed to the output layer. The neurons in the output layer also perform similar weighted summation and other calculations, and finally output the predicted value of the nonlinear compensation corresponding to the input data. In the task of predicting the nonlinear compensation of tension, the nodes in the output layer will output a specific value or a vector, which is the initial estimate of the nonlinear compensation given by the neural network based on the current input.

[0065] S33: Select an appropriate loss function based on the nature of the output task to measure the difference between the nonlinear compensation output by the network and the actual "true" compensation. For the specific compensation value of the predicted tension, the mean square error (MSE) loss function is calculated as: Where n is the number of samples, y i is the actual value, is the network prediction value.

[0066] Substitute the nonlinear compensation output by the network and the corresponding real compensation in the training data into the selected loss function to calculate the loss value after the current round of forward propagation. For example, in a batch of training samples, the loss value corresponding to the batch is calculated using the mean square error formula based on the network output and the actual value. This loss value will be used as the basis for subsequent adjustment of network parameters in the hope that the next output of the network will make the loss smaller.

[0067] S34: Based on the calculated loss value, the gradient of the connection weights and biases of each layer of neurons is calculated in reverse order from the output layer using the chain rule. For example, for a certain weight ω in the output layer, its gradient value is obtained through a series of derivative operations , this gradient indicates the degree of influence of the parameter on the loss value and the direction of change. Similarly, the gradients of the relevant parameters of other layers are calculated. These gradient information are the key to guiding how to adjust the subsequent parameters.

[0068] According to the calculated gradient, the stochastic gradient descent optimization algorithm is used to update the parameters in the network. The parameter update formula of the stochastic gradient descent optimization algorithm is: Where η is the learning rate, which determines the step size of parameter update.

[0069] By continuously adjusting parameters such as weights and biases according to the results of gradient back propagation, the network will gradually learn a more accurate mapping relationship between input and nonlinear compensation, making the subsequent output nonlinear compensation closer and closer to the actual situation.

[0070] S35: Repeat the above steps of weighted summation, loss function calculation and parameter update for multiple rounds. Each round uses a different batch of data in the training set. As the number of iterations increases, the network's learning of the nonlinear mapping relationship of the input data will become more and more in-depth, and the accuracy of the output nonlinear compensation will gradually improve.

[0071] During the training process, the validation set data is used regularly to validate the model, and the loss value or other evaluation indicators on the validation set are calculated. Observe the changes in the indicators of the validation set. If it is found that the loss value of the validation set begins to increase with the increase in the number of training rounds, corresponding measures can be taken, such as stopping training in advance and adding regularization terms to adjust the model to ensure that the model has good generalization ability and can stably output reasonable and accurate nonlinear compensation.

[0072] S36: After the above-mentioned sufficient training and verification, in actual application scenarios, for newly input tension-related data, the neural network can output the corresponding nonlinear compensation amount in the output layer according to the learned mapping relationship after the forward propagation step, and provide corresponding numerical values ​​for subsequent operations such as result fusion with the model predictive control algorithm, thereby improving the overall prediction effect.

[0073] S4: The visual inspection control center detects the coil in real time to determine whether there is an error in the winding process. If so, the winding process is stopped and an alarm is issued.

[0074] Check the appearance of the coil regularly. The visual inspection system will detect the quality of winding in real time, such as the number of turns, wire diameter, insulation performance, etc. Once a problem is found, such as wire breakage, insulation damage, wrong number of turns, whether the wires are arranged neatly, whether there is any crossing, etc., the winding machine will automatically stop and send out an alarm signal. The operator will take corresponding measures according to the alarm information, such as repairing the broken ends, replacing the insulation material, adjusting the winding parameters, etc.

[0075] S41: After the winding machine starts to wind the coil, the visual inspection system continuously collects images of the coil during the winding process at a set frequency. Each frame of the image collected is processed using the built-in image analysis algorithm. The image is first preprocessed to remove noise, enhance contrast, etc., so that features can be extracted more clearly later. Extract key features related to winding quality, such as using edge detection algorithms to identify the outline of the wire, and then determine whether the wire is arranged neatly and whether it exceeds the preset spacing range; use morphological algorithms to detect abnormal conditions such as wire breakage and knotting; use image segmentation technology to distinguish windings of different phases and check whether the connection parts are correct; S42: Based on the extracted features and pre-set quality judgment parameters, the visual inspection system judges the winding quality in real time. If the analysis results show that the current winding state meets the quality requirements, the system will continue to collect and analyze the next set of images normally; once a problem with the winding is detected, such as the wire spacing is too large and exceeds the allowable error, the wire is found to have obvious fractures, or the three-phase winding is connected incorrectly, the visual inspection system will immediately feed back the inspection results including detailed information such as the problem type and location to the winding machine control system through the communication link; S43: After receiving the winding quality problem information fed back by the visual inspection system, the winding machine control system will immediately trigger a shutdown command, and quickly stop the winding machine by controlling the power supply or driver and other related components of the motor to avoid further quality defects caused by continued winding, thereby minimizing the waste of raw materials and subsequent rework costs; S44: When the alarm signal is triggered, the winding machine control system will activate the alarm device and send out the corresponding alarm signal. The alarm signal is in the form of an audible and visual alarm, such as lighting up the warning light and emitting a continuous beep at the same time, to attract the attention of the operator, so that the operator can know in time that there is a problem with the winding and come to check and deal with it. The intensity, frequency and other parameters of the alarm signal are reasonably set according to factors such as the noise level of the actual working environment and the distance between the operator and the winding machine to ensure that abnormal information can be effectively conveyed;

[0076] S5: Place the wound coil into epoxy resin glue for degassing, ensuring that the insulating material is evenly covered on the surface of the wire and in the gap between the coil.

[0077] S51: Use the binding wire to bind the wound coil as a whole, with a spacing of 5-10 cm, to fix the coil firmly to prevent the wire from shifting during subsequent operations. Cut off the excess binding wire during the winding process to make the coil look neat. Check the appearance of the coil to ensure that there are no loose or protruding wires. For the uneven wires, use tools to gently adjust them to make them neat. Exit the winding mold and take out the wound three-phase integrated coil to avoid collision and scratching the coil surface; S52: Process the lead copper busbar into a suitable shape, remove the oxide layer and impurities on the surface to ensure its good conductivity, clean and weld the wire ends of each phase to ensure that the connection between the wire ends is firm and reliable. Use soldering equipment to weld the three lead copper busbars on each coil to the corresponding tapping terminals respectively to ensure that the welding is firm and there are no cold joints. After welding is completed, check the welding points and use a multimeter to measure the resistance to ensure a reliable connection. Insulate the welding points and lead copper busbars, and use insulating sleeves or insulating tape to wrap the connecting parts and exposed parts of the lead copper busbars to prevent short circuits or leakage. Arrange the lead copper busbars according to the design requirements to ensure that their positions are reasonable to facilitate subsequent connection with the external circuit, and maintain a certain distance between the lead copper busbars to avoid mutual interference or short circuits; S53: Mix the epoxy resin glue evenly according to a certain ratio, and then put it into a vacuum device for degassing treatment to remove bubbles in the glue liquid, ensure that the insulating material is evenly covered on the surface of the wire and in the gap between the coil, and avoid insulation weak points. Put the coil into the glue injection mold, inject the degassed epoxy resin glue into the mold through the glue injection hole, so that the glue liquid fully fills the gap of the coil and ensures that the coil is completely surrounded by the glue liquid. The glue injection process should be slow and uniform to prevent bubbles. Place the coil after glue injection in a suitable environment and cure it according to the curing requirements of the epoxy resin glue. Generally, a certain temperature and time are required, such as curing at 60-80℃ for 2-4 hours to ensure that the epoxy resin glue is fully cured and forms a solid whole. Fix the fixing plate to the coil through a bolt structure, ensure that the bolt tightening torque meets the requirements, so that the front epoxy plate and the rear epoxy plate sandwich the inner cylinder in the middle, and the rear epoxy sealing plate is fixed tightly to the rear epoxy plate. During the installation process, pay attention to the insulation treatment of the fixing plate, such as using an insulating gasket to avoid conduction with the coil or the inner cylinder.

[0078] like Figure 3 As shown, the present invention also provides a three-phase integrated coil intelligent winding and detection system for a soft start device, comprising: Electrical parameter module, used to determine the electrical parameters of coil winding; The adaptive winding module is used to adaptively control the wire tension during the coil winding process and calculate the optimal tension within the T time by establishing a tension mathematical model; A tension data acquisition unit, used to acquire the tension of the conductor using a tension sensor to obtain tension data; A tension data preprocessing unit, used for preprocessing tension data; A tension calculation unit, used to calculate the optimal tension within time T using a model predictive control algorithm; Nonlinear compensation module, used to use neural network model to estimate and correct key parameters in tension mathematical model in real time and compensate for nonlinear part; The winding quality detection module is used to detect the coil winding quality in real time using the visual detection control center to determine whether there is an error in the winding process. If so, the winding process is stopped and an alarm is issued; The post-winding processing module is used to place the wound coil into epoxy resin glue for degassing and to protect the wound coil.

[0079] The three-phase integrated coil intelligent winding and detection method for a soft starter provided by the present invention establishes a tension mathematical model for the coil winding system by using a model predictive control algorithm, calculates the optimal tension in a limited time domain, and uses a neural network model to estimate and correct key parameters in the model prediction in real time, and compensates for the nonlinear part that is difficult to accurately model by the model predictive control algorithm, thereby solving the defect that improper wire tension control affects the performance and life of the soft starter during the winding process, and the beneficial effects achieved are: MPC is good at handling constraint problems and model-based dynamic prediction, while the neural network model can deal with nonlinear and complex unknown relationships. The combination of the two gives full play to their respective advantages, improves the adaptability of the entire tension control system to complex working conditions, and can control tension more accurately. Through the assistance of the neural network model to MPC, the prediction model can be more in line with the actual coil winding system, reducing the tension control deviation caused by model errors and incompletely considered nonlinear factors, thereby ensuring the stability and uniformity of tension throughout the winding process and avoiding coil winding quality problems caused by improper tension.

[0080] By precisely controlling the winding speed and tension, the insulation and electrical performance of the coil are guaranteed. Optimizing the layout, through optimization calculation, the best layout of the winding is achieved, electromagnetic coupling and loss are reduced, and the efficiency of the equipment is improved. The high degree of automation reduces manual intervention and improves production efficiency.

[0081] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0082] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A three-phase integrated coil intelligent winding and detection method for a soft start device, characterized in that: include: S1: Determine the electrical parameters of the coil winding and prepare the winding materials; S2: During the coil winding process, the conductor tension is adaptively controlled, and the optimal tension within time T is calculated by establishing a tension mathematical model; S3: using a neural network model to estimate and correct key parameters in the tension mathematical model in real time, and to compensate for nonlinear parts; S4: The visual inspection control center detects the coil winding quality in real time and determines whether there is an error in the winding process. If so, the winding process is stopped and an alarm is issued; S5: Place the wound coil into epoxy resin glue for degassing, ensuring that the insulating material is evenly covered on the surface of the wire and in the gap between the coil.

2. The three-phase integrated coil intelligent winding and detection method for a soft start device according to claim 1 is characterized in that: In step S2, the specific steps of adaptively controlling the conductor tension are: S21: Using a tension sensor to collect the tension of the wire to obtain tension data; S22: preprocessing the tension data to obtain tension preprocessing data; S23: Use the model predictive control algorithm to establish a tension mathematical model for the coil winding system and calculate the optimal tension within time T.

3. The three-phase integrated coil intelligent winding and detection method for a soft starter device according to claim 2 is characterized in that: In step S22, the preprocessing includes data normalization, outlier removal, and periodic line inspection; Normalize the data and scale it to a specific range; The IQR method is used to calculate wild points. The degree of data dispersion is determined by calculating the quartiles of the tension training data, and whether it is a wild point is determined based on the set threshold. The steps for calculating wild points using the IQR method are as follows: Sort the data set by numerical size; Calculate the first quartile Q1. In the sorted data set, Q1 is the data point at the 25% position. Calculate the third quartile Q3. In the sorted data set, Q3 is the data point at the 75% position. Calculate IQR, IQR=Q3-Q1; After obtaining the IQR, the set threshold is used to determine whether the data point is an outlier. If so, the data point is removed. Perform periodic checks on the data to ensure that there is no duplicate data in the data set.

4. The three-phase integrated coil intelligent winding and detection method for a soft start device according to claim 2 is characterized in that: In step S23, a tension mathematical model is established for the coil winding system using a model predictive control algorithm, and the specific steps for calculating the optimal tension within time T are: S231: In the coil winding system, determining input variables and output variables; the input variables are the speed of the motor, the torque of the motor, the wire diameter and the number of turns, and the output variable is the tension of the wire; S232: Establishing a tension mathematical model based on physical principles to obtain a tension change prediction formula; S233: calculating the future tension according to the tension change prediction formula, and predicting the future tension change; S234: setting input constraints, output constraints and objective function of the wire tension during the test; using the quadratic programming algorithm to calculate the optimal output torque and speed of the motor within time T.

5. The three-phase integrated coil intelligent winding and detection method for a soft starter device according to claim 4 is characterized in that: In step S232, the tension change prediction formula is expressed as: Where F is the wire tension, T is the torque, T0 is the other resistance torque, d is the wire diameter, and N is the number of turns.

6. The three-phase integrated coil intelligent winding and detection method for a soft starter device according to claim 4 is characterized in that: In step S234, the optimal torque obtained by solving the quadratic programming is expressed as: In the formula, is the optimal control sequence at time k.

7. The three-phase integrated coil intelligent winding and detection method for a soft starter device according to claim 1 is characterized in that: In step S3, the neural network model compensates for the nonlinear part that is difficult to accurately model by the tension mathematical model predictive control algorithm, including: There is an error between the wire tension prediction value calculated by the prediction model algorithm and the wire tension actually measured; The neural network model and the model predictive control algorithm work in parallel, and the same input is input into the tension change model formula of the model predictive control algorithm and the neural network model respectively. The model predictive control algorithm outputs the prediction result according to its own linear relationship, and the neural network model outputs the corresponding nonlinear compensation amount according to the nonlinear mapping relationship it has learned. The linear relationship prediction result of the model predictive control algorithm is fused with the nonlinear relationship prediction result of the neural network to obtain a final prediction result of linear and nonlinear factors.

8. The three-phase integrated coil intelligent winding and detection method for a soft starter device according to claim 2 is characterized in that: In step S3, the specific steps of outputting the nonlinear compensation amount of the nonlinear mapping relationship of the neural network model include: S31: Select a multilayer perceptron to predict the nonlinear compensation amount of tension, and initialize the connection weights and bias items between neurons in each layer of the neural network to obtain initialization data; S32: The neurons in the hidden layer perform weighted summation on the initialization data and perform nonlinear transformation through an activation function to obtain a predicted value of a nonlinear compensation amount of the initialization data; S33: using the mean square error to calculate the loss function, and reducing the loss of the predicted value of the next nonlinear compensation amount; S34: using the chain rule to calculate the gradients of the connection weights and biases of the neurons in each layer in reverse order from the output layer, and using a stochastic gradient descent optimization algorithm to update the parameters in the neural network according to the gradients; S35: Repeat the weighted summation, loss function calculation and parameter update steps until the neural network model can stably output a reasonable and accurate nonlinear compensation amount.

9. The three-phase integrated coil intelligent winding and detection method for a soft starter device according to claim 1 is characterized in that: In step S4, the specific steps of using the visual inspection control center to monitor the winding quality in real time are as follows: S41: The visual inspection control center collects the coil image during the winding process, processes the coil image using an image analysis algorithm, and extracts image features; S42: judging that the winding quality meets the requirements according to the image features and the preset quality judgment parameters, and if so, continuing to collect and analyze the next set of images; S43: After receiving the winding quality problem information fed back by the visual inspection control center, the winding machine control center will immediately trigger a stop command to stop the winding machine; S44: After the winding machine stops running, the winding machine control center sends out an alarm.

10. A three-phase integrated coil intelligent winding and detection system for a soft starter device, which adopts the three-phase integrated coil intelligent winding and detection method for a soft starter device as claimed in any one of claims 1 to 9, characterized in that: The intelligent winding system comprises: Electrical parameter module, used to determine the electrical parameters of coil winding; The adaptive winding module is used to adaptively control the wire tension during the coil winding process and calculate the optimal tension within the T time by establishing a tension mathematical model; A tension data acquisition unit, used to acquire the tension of the conductor using a tension sensor to obtain tension data; A tension data preprocessing unit, used for preprocessing the tension data; A tension calculation unit, used to calculate the optimal tension within time T using a model predictive control algorithm; A nonlinear compensation module, used to use a neural network model to estimate and correct key parameters in the tension mathematical model in real time and to compensate for the nonlinear part; The winding quality detection module is used to detect the coil winding quality in real time using the visual detection control center to determine whether there is an error in the winding process. If so, the winding process is stopped and an alarm is issued; The post-winding processing module is used to place the wound coil into epoxy resin glue for degassing and protect the wound coil.

Images