Method, device and equipment for detecting connection of magnetic suspension motor and storage medium

Abstract

The application relates to the technical field of magnetic suspension, and provides a wiring detection method, device and equipment of a magnetic suspension motor and a storage medium. The method comprises the following steps: sending a current signal to an electromagnetic coil, controlling a rotor to float and move; collecting motion track data of the rotor; inputting the motion track data into a wiring identification model to obtain a wiring identification result, wherein the wiring identification result is used for indicating an actual wiring sequence of the electromagnetic coil; generating a wiring detection result based on the wiring identification result and a target wiring sequence, wherein the wiring detection result is used for indicating wiring correctness or wiring error. The motion track data is input into the trained wiring identification model for diagnosis, the corresponding wiring sequence of the electromagnetic coil is obtained through classification, whether the wiring is abnormal is automatically judged, manual judgment is not needed, the efficiency and accuracy of troubleshooting of the failure reason of the magnetic suspension rotor floating are improved, and the rotor floating failure problem caused by the wiring error of the electromagnetic coil is effectively solved.

Description

Technical Field

[0001] This application relates to the field of magnetic levitation technology, and in particular to a wiring detection method, device, equipment and storage medium for a magnetic levitation motor. Background Technology

[0002] Magnetic levitation bearings use magnetic force to levitate the rotor in space. Since there is no mechanical contact between the rotor and stator, there is no mechanical friction, resulting in low loss, low noise, low vibration, no need for lubrication, minimal environmental pollution, and the ability to operate at ultra-high speeds for extended periods. The working principle of a magnetic levitation bearing involves a displacement sensor monitoring the rotor's motion in real time. The collected rotor displacement signal is fed back to the control system, which then controls the current in the electromagnetic coil via a power amplifier, thereby generating a change in electromagnetic force and controlling the rotor to levitate at a given position.

[0003] In practical applications, during the manual assembly of magnetic levitation motors, the wiring sequence of the electromagnetic coils can sometimes be incorrect. For a magnetic levitation bearing system to operate stably, the correct control current must be supplied to the corresponding coils to generate the appropriate electromagnetic force on the rotor. Therefore, proper wiring of the electromagnetic coils is one of the prerequisites for the normal and stable operation of the magnetic levitation system. When the electromagnetic coil wiring is incorrect, the control of the electromagnetic force on the rotor becomes abnormal, making it impossible to guarantee the normal operation of the rotor, and may even lead to levitation failure.

[0004] In the process of assembling a magnetic levitation motor, in response to the above-mentioned problems, the relevant technologies rely heavily on on-site manual experience to troubleshoot the failure of rotor levitation caused by incorrect coil wiring. This requires multiple attempts and inferences to find the specific cause of the wiring abnormality before manual correction. This is not only slow and inefficient, but also easily affected by subjective factors and prone to misjudgment. Summary of the Invention

[0005] This application provides a wiring detection method, device, equipment, and storage medium for a magnetic levitation motor, which can improve the efficiency and accuracy of troubleshooting the causes of magnetic levitation rotor levitation failure. The technical solution is as follows:

[0006] A first aspect of this application provides a wiring detection method for a magnetic levitation motor, comprising:

[0007] A current signal is sent to the electromagnetic coil to control the rotor to lift and move;

[0008] Collect the motion trajectory data of the rotor;

[0009] The motion trajectory data is input into the wiring recognition model to obtain the wiring recognition result. The wiring recognition model is trained based on positive and negative sample data. The wiring recognition result is used to indicate the actual wiring sequence of the electromagnetic coil.

[0010] Based on the wiring identification result and the target wiring sequence, a wiring detection result is generated, which is used to indicate whether the wiring is correct or incorrect.

[0011] Optionally, before sending a current signal to the electromagnetic coil to control the rotor to float and move, the method includes:

[0012] A current signal is sent to the electromagnetic coil based on a preset rotor movement direction and a preset wiring sequence;

[0013] Collect sample motion trajectory data of the rotor to generate a rotor trajectory dataset. The rotor trajectory dataset contains positive sample motion trajectory data of the rotor under the action of a current signal with correct wiring, and negative sample motion trajectory data under the action of current signals with different types of incorrect wiring.

[0014] The wiring identification model is iteratively trained based on the rotor trajectory dataset.

[0015] Optionally, the iterative training of the wiring identification model based on the rotor trajectory dataset includes:

[0016] Initialize the weights and biases of the network in the wiring identification model;

[0017] Input the rotor trajectory dataset into the wiring recognition model and output the predicted wiring sequence;

[0018] Based on the backpropagation of the error between the predicted wiring sequence and the sample label, the model parameters of the wiring identification model are adjusted, and the sample label is used to indicate the actual wiring sequence corresponding to the sample motion trajectory data;

[0019] The next round of model training is performed based on the wire identification model with adjusted parameters until the wire identification model meets the iteration termination condition.

[0020] Optionally, inputting the rotor trajectory dataset into the wiring recognition model and outputting a predicted wiring sequence includes:

[0021] The rotor trajectory dataset is input into the input layer and hidden layer of the wiring recognition model to generate feature vectors;

[0022] The feature vector is input into the fully connected layer of the wiring recognition model, and the model is classified and the predicted wiring sequence is output through the activation function.

[0023] The step of adjusting the model parameters of the wiring identification model based on the error backpropagation between the predicted wiring sequence and the sample label includes:

[0024] The predicted wiring sequence and the sample labels are input into the loss function to generate loss function values;

[0025] In response to the wiring identification model not meeting the iteration termination condition, the model parameters are updated using the error backpropagation algorithm. The iteration termination condition is that the number of iterations reaches the target number or the loss function value is less than the loss function threshold.

[0026] Optionally, the expression of the activation function includes:

[0027]

[0028] Among them, o i and y i These are the output and probability output of the i-th neuron in the model's output layer, respectively, where K is the number of classes;

[0029] The expression for the loss function includes:

[0030]

[0031] in, For the i-th neuron of the j-th true sample vector, Let be the i-th neuron of the probability vector predicted by the j-th network, n be the total number of neurons, and M be the total number of samples.

[0032] Optionally, the iterative training of the wiring identification model based on the rotor trajectory dataset includes:

[0033] The rotor trajectory dataset is divided into a training set and a test set according to a preset ratio;

[0034] The wiring identification model is iteratively trained based on the training set in the rotor trajectory dataset.

[0035] After iteratively training the wiring identification model based on the rotor trajectory dataset, the method includes:

[0036] The wiring recognition model was tested based on the test set, and the model recognition accuracy was calculated.

[0037] When the model's recognition accuracy reaches the accuracy threshold, the model training ends.

[0038] In response to the fact that the model recognition accuracy does not reach the accuracy threshold, the model parameters of the wiring recognition model are adjusted and iterative training continues.

[0039] Optionally, after generating the wiring detection result based on the wiring identification result and the target wiring sequence, the method further includes:

[0040] In response to the wiring detection result indicating a wiring error, the wiring pins of the electromagnetic coil are updated based on the target wiring sequence and the actual wiring sequence.

[0041] A second aspect of this application provides a wiring detection device for a magnetic levitation motor, comprising:

[0042] The transmitting module is used to send current signals to the electromagnetic coil to control the rotor to lift and move.

[0043] The acquisition module is used to acquire the motion trajectory data of the rotor;

[0044] The identification module is used to input the motion trajectory data into the wiring identification model to obtain the wiring identification result. The wiring identification model is trained based on positive and negative sample data. The wiring identification result is used to indicate the actual wiring sequence of the electromagnetic coil.

[0045] The detection module is used to generate a wiring detection result based on the wiring identification result and the target wiring sequence. The wiring detection result is used to indicate whether the wiring is correct or incorrect.

[0046] A third aspect of this application provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the method described above.

[0047] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described method.

[0048] A fifth aspect of this application provides a computer program product comprising a computer program or instructions that, when executed by a processor, implement the steps of the above-described method.

[0049] The technical solution provided in this application includes at least the following beneficial effects:

[0050] The wiring detection method, device, equipment, and storage medium for magnetic levitation motors provided in this application control the rotor movement by sending a specific current signal and collecting the rotor's motion trajectory data. The motion trajectory data is then input into a trained wiring recognition model for diagnosis, classifying the corresponding electromagnetic coil wiring sequence and automatically determining whether there are any wiring abnormalities. This eliminates the need for manual judgment, improves the efficiency and accuracy of troubleshooting the causes of magnetic levitation rotor levitation failure, and effectively solves the problem of rotor levitation failure caused by electromagnetic coil wiring errors. Attached Figure Description

[0051] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0052] Figure 1 This is a schematic diagram of an application scenario of the wiring detection method for a magnetic levitation motor provided in the embodiments of this application;

[0053] Figure 2 This is a flowchart of some embodiments of a wiring detection method for a magnetic levitation motor provided in this application;

[0054] Figure 3 This is a trajectory diagram of a rotor performing a preset motion under normal wiring conditions, according to an embodiment of this application.

[0055] Figure 4 This is a trajectory diagram of a rotor performing a preset motion under an incorrect wiring condition, according to an embodiment of this application.

[0056] Figure 5 These are flowcharts of some other embodiments of a wiring detection method for a magnetic levitation motor provided in the embodiments of this application;

[0057] Figure 6 This is a model framework diagram of a wiring detection method for a magnetic levitation motor provided according to an embodiment of this application;

[0058] Figure 7 This is a model training flowchart of a wiring detection method for a magnetic levitation motor according to an embodiment of this application;

[0059] Figure 8 This is a structural block diagram of a wiring detection device for a magnetic levitation motor provided according to an embodiment of this application;

[0060] Figure 9 This is a structural block diagram of an electronic device provided according to an embodiment of this application. Detailed Implementation

[0061] Embodiments of this application will now be described in more detail with reference to the accompanying drawings. While some embodiments of this application are shown in the drawings, it should be understood that this application can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this application. It should be understood that the drawings and embodiments of this application are for illustrative purposes only and are not intended to limit the scope of protection of this application.

[0062] It should also be noted that, for ease of description, only the parts relevant to this application are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this application can be combined with each other.

[0063] It should be noted that the concepts of "first" and "second" mentioned in this application are only used to distinguish different systems, devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these systems, devices, modules or units.

[0064] It should be noted that the terms "a" and "a plurality of" used in this application are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0065] The names of the messages or information exchanged between multiple devices in the embodiments of this application are for illustrative purposes only and are not intended to limit the scope of these messages or information.

[0066] Figure 1 This is a schematic diagram of an application scenario provided according to some embodiments of this application.

[0067] exist Figure 1 In application scenarios, the magnetic levitation stator is equipped with multiple control coils, as shown in the attached diagram. Figure 1 The schematic diagram of the radial magnetic levitation bearing is shown, such as electromagnetic coil 1 in the X direction, electromagnetic coil 2 in the X direction, electromagnetic coil 1 in the Y direction, and electromagnetic coil 2 in the Y direction. The control system generates corresponding electromagnetic force by sending current signals to the corresponding electromagnetic coils to control the rotor to levitate. Incorrect wiring of the electromagnetic coils will cause the rotor to fail to levitate normally.

[0068] The control system can be either hardware or software, deployed within electronic devices. When the control system is hardware, it can be implemented as a distributed cluster of multiple servers or terminal devices, or as a single server or terminal device. When the control system is software, it can be installed on the hardware devices listed above. It can be implemented as, for example, multiple software programs or software modules used to provide distributed services, or as a single software program or software module. No specific limitations are imposed here.

[0069] Continue to refer to Figure 2 The diagram illustrates a flowchart of a wiring detection method for a magnetic levitation motor provided in an exemplary embodiment of this application. The method includes the following steps:

[0070] Step 201: Send a current signal to the electromagnetic coil to control the rotor to lift and move.

[0071] In one possible implementation, the detection phase begins after the coil wiring is completed. The electronic device stores a target wiring sequence, i.e., the correct wiring sequence. Based on a preset motion trajectory, the device sends a current signal to the electromagnetic coil according to the target wiring sequence. For example, the device generates a current signal that, under the target wiring sequence, can control the rotor to perform counterclockwise circular motion.

[0072] Step 202: Collect the rotor's motion trajectory data.

[0073] In one possible implementation, the device collects the motion trajectory data of the magnetically levitated rotor in real time using a displacement sensor. Specifically, the device collects the rotor's displacement data multiple times at a preset frequency to obtain the motion trajectory data.

[0074] If the wiring is correct (i.e., the actual wiring sequence matches the target wiring sequence), then the motion trajectory formed by the rotor's motion trajectory data should be consistent with the preset motion trajectory (e.g., Figure 2 The diagram shows a counter-clockwise circular motion. If the wiring is incorrect, the rotor's trajectory will be a different shape (e.g., ...). Figure 3 (The irregular motion trajectory shown). Therefore, this embodiment controls the rotor's movement and simultaneously collects the rotor's motion trajectory data to determine whether its motion trajectory is a preset trajectory, thus eliminating coil wiring faults.

[0075] Step 203: Input the motion trajectory data into the wiring recognition model to obtain the wiring recognition result. The wiring recognition model is trained based on positive and negative sample data. The wiring recognition result is used to indicate the actual wiring sequence of the electromagnetic coil.

[0076] Electronic devices are equipped with wiring recognition models. The device inputs motion trajectory data into the wiring recognition model to obtain the wiring recognition results. Specifically, the model output includes the probabilities of various wiring sequences, and the electronic device identifies the wiring sequence with the highest probability as the actual wiring sequence.

[0077] In one possible implementation, the wiring identification model is pre-trained based on positive and negative sample data. The positive sample data consists of sample motion trajectory data of the rotor under correct wiring sequence, while the negative sample data consists of sample motion trajectory data of the rotor under various incorrect wiring sequences.

[0078] Step 204: Based on the wiring identification result and the target wiring sequence, generate the wiring detection result, which is used to indicate whether the wiring is correct or incorrect.

[0079] The electronic device compares the actual wiring sequence indicated by the wiring identification result with the target wiring sequence. If they match, the wiring is considered correct; otherwise, the wiring is considered incorrect.

[0080] In summary, the wiring detection method for magnetic levitation motors provided in this application controls the rotor movement by sending a specific current signal and collecting the rotor's motion trajectory data. The motion trajectory data is then input into a trained wiring recognition model for diagnosis, classifying the corresponding electromagnetic coil wiring sequence and automatically determining whether there are any wiring abnormalities. This eliminates the need for manual judgment, improves the efficiency and accuracy of troubleshooting the causes of magnetic levitation rotor levitation failure, and effectively solves the problem of rotor levitation failure caused by electromagnetic coil wiring errors.

[0081] Optionally, if an electronic device detects a wiring error, it can automatically correct the wiring sequence through a program, without requiring manual disassembly. Following step 204 above, the method of this embodiment further includes the following steps:

[0082] In response to a wiring error detection result, the wiring pins of the electromagnetic coil are updated based on the target wiring sequence and the actual wiring sequence.

[0083] like Figure 1 As shown, the control system obtains the corresponding input signal through the interface connected to the coil leads. After identifying the current lead sequence of the coil, if there is an error, it can be automatically corrected by the program. The program can specify the coil input signal corresponding to each interface, complete the correction of the wiring sequence, and change the corresponding wiring pins without the need for manual disconnection and insertion of leads.

[0084] See also Figure 5 Before the equipment uses the wiring identification model to troubleshoot wiring faults, it is necessary to first train the model using sample data. Figure 5 The flowchart of the model training phase is shown. Before step 201 above, the method also includes the following steps:

[0085] Step 501: Send a current signal to the electromagnetic coil based on the preset rotor movement direction and preset wiring sequence.

[0086] The first stage involves collecting the rotor trajectory data. The electronic device's control system delivers current to each coil according to a specific control strategy, causing the rotor to perform a specific motion, such as making the rotor move in a circular motion similar to clockwise or counterclockwise. Simultaneously, displacement sensors collect the motion trajectory data of the magnetically levitated rotor. Here, we will use counterclockwise as an example for illustration:

[0087] like Figure 3 As shown, a reference xy coordinate system is established with the stator geometric center C as the origin. The rotor initially rests on the protective bearing. A control current is applied to the coil to make the rotor move in a similar circular motion in a counterclockwise direction.

[0088] like Figure 3The diagram shows the normal trajectory of the magnetic levitation rotor when the X-direction electromagnetic coil wiring 1, X-direction electromagnetic coil wiring 2, Y-direction electromagnetic coil wiring 1, and Y-direction electromagnetic coil wiring 2 are properly connected, and controlled by a certain control strategy to perform a circular motion. Figure 4 This diagram illustrates the abnormal trajectory of a magnetically levitated rotor when there is a wiring error in the X-direction electromagnetic coil wiring 1, X-direction electromagnetic coil wiring 2, Y-direction electromagnetic coil wiring 1, and Y-direction electromagnetic coil wiring 2, and the rotor is controlled to rotate according to the original control strategy.

[0089] Step 502: Collect sample motion trajectory data of the rotor to generate a rotor trajectory dataset.

[0090] The rotor trajectory dataset contains positive sample motion trajectory data of the rotor under the action of a current signal with correct wiring, and negative sample motion trajectory data under the action of current signals with different types of incorrect wiring.

[0091] To illustrate, the rotor is driven to rotate once every time interval T, and the current rotor position x is recorded once. n y n The rotor motion trajectory matrix P = [x1, y1, x2, y2, ..., x] is obtained. n y n ] T Under various coil wiring conditions, multiple sets of rotor motion trajectory samples were collected and labeled accordingly. A total of M motion trajectory samples were collected, denoted as X = {P1, P2, ..., P...}. M}

[0092] Step 503: Iteratively train the wiring identification model based on the rotor trajectory dataset.

[0093] In one possible implementation, the wiring identification model is constructed using a one-dimensional convolutional neural network (CNN). An illustrative example of the constructed 1D CNN network model is attached. Figure 6 As shown, the system mainly consists of an input layer, a one-dimensional convolutional layer, a one-dimensional pooling layer, a fully connected layer, and an output layer. The hidden layers consist of convolutional and pooling layers. The convolutional layers are responsible for feature extraction from the input data, while the pooling layers reduce the dimensionality of the convolutional layer's output, decreasing the data size. The convolutional and pooling layers are alternately connected to jointly complete feature mining of the original rotor trajectory data. The data is then further extracted by the fully connected layer. Finally, the output of the fully connected layer is connected to a softmax classifier to output the classification result, thus achieving the task of classifying the rotor trajectory data.

[0094] To enable a CNN model to classify and learn, the network model needs to be trained. Training includes two stages: forward propagation and back propagation. Forward propagation mainly inputs the sample training set into the network to obtain the output. Back propagation mainly compares the actual output with the ideal output to obtain the error between the two, and then calculates the derivative to adjust the network parameters until the iteration termination condition is met. The method in this embodiment of the application identifies the coil wiring state by inputting the rotor trajectory into a 1D CNN model. The overall model training flowchart is attached. Figure 7 As shown, it includes three stages: collection of rotor trajectory dataset and model training, and model testing.

[0095] Specifically, step 503 includes the following steps:

[0096] Step 503a: Initialize the weights and biases of the network in the wiring identification model.

[0097] Optionally, the device initializes the weights and biases of each layer of the 1DCNN model, and sets the learning rate and loss function for training.

[0098] Step 503b: Input the rotor trajectory dataset into the wiring recognition model and output the predicted wiring sequence.

[0099] A forward propagation is performed, extracting features from the rotor trajectory training set from the input layer to the hidden layers (convolutional-pooling layers), and then classifying them in the fully connected layer to output the predicted wiring sequence. The specific process is as follows:

[0100] Step 1: Input the rotor trajectory dataset into the input layer and hidden layer of the wiring recognition model to generate feature vectors.

[0101] Step 2: Input the feature vector into the fully connected layer of the wiring recognition model, classify it through the activation function, and output the predicted wiring sequence.

[0102] Indicatively, the model in this embodiment uses the softmax activation function, the expression of which includes:

[0103]

[0104] Among them, o i and y i and are the output and probability output of the i-th neuron in the model's output layer, respectively, where K is the number of categories.

[0105] Step 503c: Based on the backpropagation of the error between the predicted wiring sequence and the sample label, adjust the model parameters of the wiring identification model. The sample label is used to indicate the actual wiring sequence corresponding to the sample motion trajectory data.

[0106] Step 503d: Perform the next round of model training based on the wire identification model with adjusted parameters until the wire identification model meets the iteration termination condition.

[0107] After the current training model outputs the predicted wiring sequence, backpropagation is required. The backpropagation process is the reverse of the forward propagation process. It involves comparing the predicted classification result of each sample with the actual labeled classification result, calculating the model loss, and updating the network parameters (weights and biases) using the error backpropagation algorithm. This process is repeated until the wiring recognition model meets the iteration termination condition. The specific process is as follows:

[0108] Step 3: Input the predicted wiring sequence and sample labels into the loss function to generate the loss function value.

[0109] Step 4: In response to the wiring identification model not meeting the iteration termination condition, the error backpropagation algorithm is used to update the model parameters. The iteration termination condition is that the number of iterations reaches the target number or the loss function value is less than the loss function threshold.

[0110] To illustrate, the loss function used is the cross-entropy loss function, and the expression for the loss function includes:

[0111]

[0112] in, For the i-th neuron of the j-th true sample vector, Let be the i-th neuron of the probability vector predicted by the j-th network, n be the total number of neurons, and M be the total number of samples.

[0113] After the model training is completed, model testing is required. In one possible implementation, the device divides the dataset into a training set and a test set according to a preset ratio before model training, for use in model training and testing respectively. Step 503 includes the following steps:

[0114] Step 5: Divide the rotor trajectory dataset into a training set and a test set according to a preset ratio.

[0115] Step 6: Iteratively train the wiring recognition model based on the training set in the rotor trajectory dataset.

[0116] The method provided in this application embodiment further includes the following steps:

[0117] Step 7: Test the wiring recognition model based on the test set and calculate the model recognition accuracy.

[0118] Step 8: In response to the model recognition accuracy reaching the accuracy threshold, end model training.

[0119] Step 9: In response to the fact that the model recognition accuracy has not reached the accuracy threshold, adjust the model parameters of the wiring recognition model and continue iterative training.

[0120] The rotor trajectory test dataset is input into the trained 1DCNN network model. Through forward propagation, classification and diagnostic results are obtained, thereby identifying the coil wiring status. Multiple sets of recognition results are obtained through multiple predictions, and the model's recognition accuracy is calculated based on the corresponding labels. If the accuracy reaches the standard, model training stops, and the model can be used for practical applications. If the standard is not reached, the parameters are readjusted, and training continues.

[0121] In this embodiment, a dataset X of rotor displacement trajectories of the magnetic levitation rotor under different electromagnetic coil wiring states is collected. A one-dimensional convolutional neural network model is established and trained. The trained model is used to diagnose the input rotor displacement trajectory and classify the corresponding electromagnetic coil wiring sequence state. If the wiring is abnormal, the program automatically corrects it without manual disassembly, saving time and avoiding electromagnetic coil wiring errors. This improves the efficiency and accuracy of troubleshooting the cause of magnetic levitation rotor levitation failure and effectively solves the problem of rotor levitation failure caused by electromagnetic coil wiring errors.

[0122] The following are embodiments of the apparatus described in this application, which can be used to execute the embodiments of the method described in this application. For details not disclosed in the apparatus embodiments of this application, please refer to the embodiments of the method described in this application.

[0123] Further reference Figure 8 As an implementation of the methods described in the above figures, this application provides some embodiments of a wiring detection device for a magnetic levitation motor, which correspond to the method embodiments described above.

[0124] like Figure 8 As shown, the wiring detection device for a magnetic levitation motor in some embodiments includes:

[0125] The transmitting module 801 is used to send current signals to the electromagnetic coil to control the rotor to lift and move.

[0126] Acquisition module 802 is used to acquire the motion trajectory data of the rotor;

[0127] The identification module 803 is used to input the motion trajectory data into the wiring identification model to obtain the wiring identification result. The wiring identification model is trained based on positive and negative sample data. The wiring identification result is used to indicate the actual wiring sequence of the electromagnetic coil.

[0128] The detection module 804 is used to generate a wiring detection result based on the wiring identification result and the target wiring sequence. The wiring detection result is used to indicate whether the wiring is correct or incorrect.

[0129] Optionally, the transmitting module 801 is further configured to transmit a current signal to the electromagnetic coil based on a preset rotor movement direction and a preset wiring sequence;

[0130] The acquisition module 802 is also used to acquire sample motion trajectory data of the rotor and generate a rotor trajectory dataset. The rotor trajectory dataset includes positive sample motion trajectory data of the rotor under the action of a current signal with correct wiring, and negative sample motion trajectory data under the action of current signals with different types of incorrect wiring.

[0131] The device also includes a training module for iteratively training the wiring identification model based on the rotor trajectory dataset.

[0132] Optionally, the training module is further configured to:

[0133] Initialize the weights and biases of the network in the wiring identification model;

[0134] Input the rotor trajectory dataset into the wiring recognition model and output the predicted wiring sequence;

[0135] Based on the backpropagation of the error between the predicted wiring sequence and the sample label, the model parameters of the wiring identification model are adjusted, and the sample label is used to indicate the actual wiring sequence corresponding to the sample motion trajectory data;

[0136] The next round of model training is performed based on the wire identification model with adjusted parameters until the wire identification model meets the iteration termination condition.

[0137] Optionally, the training module is further configured to:

[0138] The rotor trajectory dataset is input into the input layer and hidden layer of the wiring recognition model to generate feature vectors;

[0139] The feature vector is input into the fully connected layer of the wiring recognition model, and the model is classified and the predicted wiring sequence is output through the activation function.

[0140] The predicted wiring sequence and the sample labels are input into the loss function to generate loss function values;

[0141] In response to the wiring identification model not meeting the iteration termination condition, the model parameters are updated using the error backpropagation algorithm. The iteration termination condition is that the number of iterations reaches the target number or the loss function value is less than the loss function threshold.

[0142] Optionally, the expression of the activation function includes:

[0143]

[0144] Among them, o i and y i These are the output and probability output of the i-th neuron in the model's output layer, respectively, where K is the number of classes;

[0145] The expression for the loss function includes:

[0146]

[0147] in, For the i-th neuron of the j-th true sample vector, Let be the i-th neuron of the probability vector predicted by the j-th network, n be the total number of neurons, and M be the total number of samples.

[0148] Optionally, the training module is further configured to:

[0149] The rotor trajectory dataset is divided into a training set and a test set according to a preset ratio;

[0150] The wiring identification model is iteratively trained based on the training set in the rotor trajectory dataset.

[0151] The device also includes a testing module for:

[0152] The wiring recognition model was tested based on the test set, and the model recognition accuracy was calculated.

[0153] When the model's recognition accuracy reaches the accuracy threshold, the model training ends.

[0154] In response to the fact that the model recognition accuracy does not reach the accuracy threshold, the model parameters of the wiring recognition model are adjusted and iterative training continues.

[0155] It is understandable that the modules described in the device correspond to the various steps in the method. Therefore, the operations, features, and beneficial effects described above for the method also apply to the device and the modules it contains, and will not be repeated here.

[0156] like Figure 9 As shown, electronic device 900 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 901, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 902 or a program loaded from storage device 908 into random access memory (RAM) 903. RAM 903 also stores various programs and data required for the operation of electronic device 900. Processing device 901, ROM 902, and RAM 903 are interconnected via bus 904. Input / output (I / O) interface 905 is also connected to bus 904.

[0157] Typically, the following devices can be connected to I / O interface 905: input devices 906 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 907 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 908 including, for example, magnetic tapes, hard disks, etc.; and communication devices 909. Communication device 909 allows electronic device 900 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 9 An electronic device 900 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 9 Each box shown can represent a device or multiple devices as needed.

[0158] In particular, according to some embodiments of this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this application include a computer program product comprising a computer program or instructions carried on a computer-readable medium, the computer program or instructions containing program code for performing the methods shown in the flowcharts. It should be noted that the computer-readable medium described above in some embodiments of this application can be a computer-readable signal medium or a computer-readable storage medium, or any combination of the two. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof.

[0159] The above description is merely a selection of preferred embodiments of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this application.

Claims

1. A wiring detection method for a magnetic levitation motor, characterized in that, include: A current signal is sent to the electromagnetic coil to control the rotor to lift and move; Collect the motion trajectory data of the rotor; The motion trajectory data is input into the wiring recognition model to obtain the wiring recognition result. The wiring recognition model is trained based on positive and negative sample data. The wiring recognition result is used to indicate the actual wiring sequence of the electromagnetic coil. The wiring recognition model is a one-dimensional convolutional neural network. Based on the wiring identification result and the target wiring sequence, a wiring detection result is generated, which is used to indicate whether the wiring is correct or incorrect. Before sending a current signal to the electromagnetic coil to control the rotor to float and move, the method includes: A current signal is sent to the electromagnetic coil based on a preset rotor movement direction and a preset wiring sequence; The position coordinates of the rotor are collected once at a preset time interval to generate the motion trajectory matrix of the rotor in the current wiring state. Multiple sets of motion trajectory matrices collected under various different wiring states are used to generate a rotor trajectory dataset. The rotor trajectory dataset includes positive sample motion trajectory data of the rotor under the action of the current signal with correct wiring, and negative sample motion trajectory data under the action of the current signal with different types of incorrect wiring. The wiring identification model is iteratively trained based on the rotor trajectory dataset.

2. The method according to claim 1, characterized in that, The iterative training of the wiring identification model based on the rotor trajectory dataset includes: Initialize the weights and biases of the network in the wiring identification model; Input the rotor trajectory dataset into the wiring recognition model and output the predicted wiring sequence; Based on the backpropagation of the error between the predicted wiring sequence and the sample label, the model parameters of the wiring identification model are adjusted, and the sample label is used to indicate the actual wiring sequence corresponding to the sample motion trajectory data; The next round of model training is performed based on the wire identification model with adjusted parameters until the wire identification model meets the iteration termination condition.

3. The method according to claim 2, characterized in that, The step of inputting the rotor trajectory dataset into the wiring recognition model and outputting a predicted wiring sequence includes: The rotor trajectory dataset is input into the input layer and hidden layer of the wiring recognition model to generate feature vectors; The feature vector is input into the fully connected layer of the wiring recognition model, and the model is classified and the predicted wiring sequence is output through the activation function. The step of adjusting the model parameters of the wiring identification model based on the error backpropagation between the predicted wiring sequence and the sample label includes: The predicted wiring sequence and the sample labels are input into the loss function to generate loss function values; In response to the wiring identification model not meeting the iteration termination condition, the model parameters are updated using the error backpropagation algorithm. The iteration termination condition is that the number of iterations reaches the target number or the loss function value is less than the loss function threshold.

4. The method according to claim 3, characterized in that, The expression for the activation function includes: Among them, o i and y i These are the output and probability output of the i-th neuron in the model's output layer, respectively, where K is the number of classes; The expression for the loss function includes: in, For the i-th neuron of the j-th true sample vector, Let be the i-th neuron of the probability vector predicted by the j-th network, n be the total number of neurons, and M be the total number of samples.

5. The method according to any one of claims 1 to 4, characterized in that, The iterative training of the wiring identification model based on the rotor trajectory dataset includes: The rotor trajectory dataset is divided into a training set and a test set according to a preset ratio; The wiring identification model is iteratively trained based on the training set in the rotor trajectory dataset. After iteratively training the wiring identification model based on the rotor trajectory dataset, the method includes: The wiring recognition model was tested based on the test set, and the model recognition accuracy was calculated. When the model's recognition accuracy reaches the accuracy threshold, the model training ends. In response to the fact that the model recognition accuracy does not reach the accuracy threshold, the model parameters of the wiring recognition model are adjusted and iterative training continues.

6. The method according to any one of claims 1 to 4, characterized in that, After generating the wiring detection result based on the wiring identification result and the target wiring sequence, the method further includes: In response to the wiring detection result indicating a wiring error, the wiring pins of the electromagnetic coil are updated based on the target wiring sequence and the actual wiring sequence.

7. A wiring detection device for a magnetic levitation motor, characterized in that, include: The transmitting module is used to send current signals to the electromagnetic coil to control the rotor to lift and move. The acquisition module is used to acquire the motion trajectory data of the rotor; The identification module is used to input the motion trajectory data into the wiring identification model to obtain the wiring identification result. The wiring identification model is trained based on positive and negative sample data. The wiring identification result is used to indicate the actual wiring sequence of the electromagnetic coil. The wiring identification model is a one-dimensional convolutional neural network. The detection module is used to generate a wiring detection result based on the wiring identification result and the target wiring sequence, the wiring detection result being used to indicate whether the wiring is correct or incorrect; The training module, before sending a current signal to the electromagnetic coil to control the rotor to lift and move, is used for: A current signal is sent to the electromagnetic coil based on a preset rotor movement direction and a preset wiring sequence; The position coordinates of the rotor are collected once at a preset time interval to generate the motion trajectory matrix of the rotor in the current wiring state. Multiple sets of motion trajectory matrices collected under various different wiring states are used to generate a rotor trajectory dataset. The rotor trajectory dataset includes positive sample motion trajectory data of the rotor under the action of the current signal with correct wiring, and negative sample motion trajectory data under the action of the current signal with different types of incorrect wiring. The wiring identification model is iteratively trained based on the rotor trajectory dataset.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 6.

Images