Sectional type PMLSM stroke self-detection method

Abstract

The invention discloses a sectional type PMLSM stroke self-detection method, relates to the technical field of PMLSM stroke detection, and solves the technical problems of low reliability, high cost, insufficient precision and robustness and the like in a complex environment in the prior art. The method comprises the following steps: acquiring electromagnetic data of the permanent magnet synchronous linear motor during operation; building a finite element model of the permanent magnet synchronous linear motor based on finite element simulation software; sampling point marking is carried out on a running orbit of the permanent magnet synchronous linear motor rotor in the finite element model to obtain position point data; calculating electromagnetic data corresponding to a sampling point in each piece of position point data through a finite element simulation algorithm, and training an artificial intelligence model based on the position point data and the corresponding electromagnetic data to obtain a rotor position recognition model; identifying the electromagnetic data acquired in real time based on a rotor position identification model to obtain an identification result; and whether early warning needs to be carried out is judged based on the identification result. The technical problem is solved.

Description

Technical Field

[0001] The invention belongs to the field of PMLSM stroke detection, in particular to a segmented PMLSM stroke self-detection method. Background Art

[0002] Permanent magnet linear synchronous motor (PMLSM) is widely used in precision manufacturing, automated production lines, high-speed transportation and other fields due to its high efficiency, high precision and good dynamic response characteristics; however, the mover position detection of PMLSM has always been one of the key technologies.

[0003] The existing PMLSM stroke self-detection methods mainly rely on position sensors and methods based on single electromagnetic parameters or mathematical models. Although these methods can provide certain position information, they have problems such as low reliability, high cost, difficulty in installation and maintenance, insufficient accuracy and robustness in complex environments. At the same time, in practical applications, field data collection requires complex sensor arrangements and data acquisition systems, which increases the complexity and cost of the system. In addition, the data collected on site may be interfered by environmental noise, which affects the data quality and the effect of model training.

[0004] Therefore, the present invention proposes a segmented PMLSM stroke self-detection method to solve the above-mentioned problem. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a segmented PMLSM stroke self-detection method, which is used to solve the problems that the existing PMLSM stroke self-detection methods mainly rely on position sensors and methods based on single electromagnetic parameters or mathematical models. Although these methods can provide certain position information, they have low reliability, high cost, difficulty in installation and maintenance, insufficient accuracy and robustness in complex environments, etc.; at the same time, in practical applications, field data collection requires complex sensor arrangements and data acquisition systems, which increases the complexity and cost of the system, and the data collected on site may be interfered with by environmental noise, resulting in technical problems that affect data quality and the effect of model training.

[0006] To achieve the above object, the first aspect of the present invention provides a segmented PMLSM stroke self-detection method, comprising:

[0007] Step 1: Obtain electromagnetic data of the permanent magnet synchronous linear motor during operation;

[0008] Step 2: Construct a finite element model of the permanent magnet synchronous linear motor based on finite element simulation software;

[0009] Step 3: Mark sampling points on the running track of the permanent magnet synchronous linear motor mover in the finite element model to obtain position point data;

[0010] Step 4: Calculate the electromagnetic data corresponding to the sampling point in each position point data through the finite element simulation algorithm, and train the artificial intelligence model based on the position point data and the corresponding electromagnetic data to obtain the mover position recognition model;

[0011] Step 5: Identify the electromagnetic data acquired in real time through the mover position recognition model to obtain the recognition result; determine whether an early warning is needed based on the recognition result.

[0012] Preferably, the step of obtaining electromagnetic data of the permanent magnet linear synchronous motor during operation includes:

[0013] The electromagnetic data of the permanent magnet synchronous linear motor during operation is collected by a data acquisition device; wherein the electromagnetic data includes: current, voltage, back electromotive force and air gap flux density.

[0014] It should be noted that the position of the mover is predicted by the air gap flux density because when the mover moves, the relative position between the primary and secondary changes, resulting in a change in the distribution of the air gap flux density, and the different degrees of magnetic saturation at different positions result in different sizes of the air gap flux density;

[0015] The position of the mover can be predicted by back electromotive force because when the mover moves, the magnetic flux in the winding changes. According to Faraday's law of electromagnetic induction, the back electromotive force is proportional to the rate of change of the magnetic flux, and the back electromotive force is proportional to the speed of the mover. The change in speed will affect the magnitude of the back electromotive force.

[0016] The position of the mover is predicted by current and voltage because the mover may encounter different loads at different positions, resulting in current changes. The magnetic field strength at different positions is different, affecting the current in the winding; the characteristics of the power supply (such as voltage fluctuations) will affect the voltage of the motor, and changes in back electromotive force will also affect the voltage at both ends of the winding.

[0017] Preferably, the finite element model of the permanent magnet synchronous linear motor is constructed based on finite element simulation software, including:

[0018] A1: Build a 2D or 3D model of the permanent magnet synchronous linear motor using finite element simulation software; the finite element simulation software includes: ANSYS, COMSOL or FEMM;

[0019] A2: Define the material properties in the model; the material properties include: primary core, secondary plate, permanent magnet, air gap, armature core, armature winding and air medium;

[0020] A3: Define the boundary conditions and excitation sources of the model.

[0021] Preferably, the marking of sampling points on the running track of the permanent magnet linear synchronous motor mover in the finite element model includes:

[0022] B1: extracting the running track length range of the mover of the target permanent magnet synchronous linear motor; wherein the target permanent magnet synchronous linear motor refers to the permanent magnet synchronous linear motor to be detected;

[0023] B2: Call the lhsdesign function in MATLAB to perform Latin hypercube sampling within the orbit length range to generate sampling points;

[0024] B3: Based on the sampling points, the position of each sampling point is marked in the finite element model to obtain the position point data; wherein the position point data includes: the specific position of the sampling point on the moving track and the distance between the sampling point and the initial moving position.

[0025] Preferably, the training of the artificial intelligence model based on the location point data and the corresponding electromagnetic data includes:

[0026] Integrate the electromagnetic data corresponding to the sampling points in the position point data into standard input data, and integrate the position point data into standard output data;

[0027] The artificial intelligence model is trained based on standard input data and standard output data to obtain a mover position recognition model; wherein the artificial intelligence model includes: a convolutional neural network or a deep belief network.

[0028] Preferably, the identifying of electromagnetic data acquired in real time by using a mover position identification model includes:

[0029] The electromagnetic data of the permanent magnet synchronous linear motor during operation, which is collected in real time, is input into the mover position identification model to obtain an identification result; wherein the identification result includes: the specific position of the actual mover on the mover running track in the finite element model and the distance between the actual mover and the initial mover position.

[0030] Preferably, judging whether an early warning is needed based on the recognition result includes:

[0031] Extract the distance between the actual mover and the initial mover position in the recognition result and mark it as the travel distance;

[0032] Determine whether the travel distance exceeds the preset travel distance range within the same time period; if yes, generate an early warning message and send it to the client; if no, continue to detect and judge.

[0033] It should be noted that the preset travel distance range is set according to the error requirement of the mover movement of the actual permanent magnet synchronous linear motor during operation.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] The existing PMLSM stroke self-detection methods mainly rely on position sensors and methods based on single electromagnetic parameters or mathematical models. Although these methods can provide certain position information, they have the problems of low reliability, high cost, difficulty in installation and maintenance, insufficient accuracy and robustness in complex environments; at the same time, in practical applications, field data collection requires complex sensor arrangements and data acquisition systems, which increases the complexity and cost of the system, and the field collected data may be interfered by environmental noise, resulting in problems such as affecting data quality and the effect of model training; the present invention generates high-quality electromagnetic data by constructing a finite element model of PMLSM in finite element simulation software, thereby avoiding the field data collection. The complexity and uncertainty of the set are solved; first, the multi-parameter electromagnetic data of the motor during operation is obtained through the data acquisition equipment, and the sampling points are marked in the finite element model to calculate the electromagnetic data of each sampling point; then, the artificial intelligence models such as convolutional neural networks or deep belief networks are used to train these data to obtain the mover position recognition model; finally, the mover position recognition model is used to identify the electromagnetic data collected in real time, determine whether the mover position exceeds the preset travel distance range, and generate warning information, which not only avoids the problems of low reliability, high cost, installation and maintenance difficulties brought by the method relying on position sensors and single electromagnetic parameters or mathematical models, but also improves the accuracy and robustness of position detection. At the same time, by using artificial intelligence models for data processing and recognition, the complexity requirements of sensor layout and data acquisition system for field data collection are effectively reduced, and the impact of environmental noise on data quality and model training effect is reduced, thereby achieving more efficient, accurate and reliable PMLSM travel self-detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0037] Figure 1 A schematic diagram of the method steps of an embodiment of the present invention;

[0038] Figure 2 Schematic diagram of sampling point marking according to an embodiment of the present invention. DETAILED DESCRIPTION

[0039] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] See also Figure 1 The first aspect of the present invention provides a segmented PMLSM stroke self-detection method, comprising:

[0041] Step 1: Obtain electromagnetic data of the permanent magnet synchronous linear motor during operation;

[0042] Step 2: Construct a finite element model of the permanent magnet synchronous linear motor based on finite element simulation software;

[0043] Step 3: Mark sampling points on the running track of the permanent magnet synchronous linear motor mover in the finite element model to obtain position point data;

[0044] Step 4: Calculate the electromagnetic data corresponding to the sampling point in each position point data through the finite element simulation algorithm, and train the artificial intelligence model based on the position point data and the corresponding electromagnetic data to obtain the mover position recognition model;

[0045] Step 5: Identify the electromagnetic data acquired in real time through the mover position recognition model to obtain the recognition result; determine whether an early warning is needed based on the recognition result.

[0046] Obtain electromagnetic data of permanent magnet synchronous linear motor during operation, including:

[0047] The electromagnetic data of the permanent magnet synchronous linear motor during operation is collected by a data acquisition device; wherein the electromagnetic data includes: current, voltage, back electromotive force and air gap flux density.

[0048] The finite element model of the permanent magnet synchronous linear motor is constructed based on the finite element simulation software, including:

[0049] A1: Build a 2D or 3D model of the permanent magnet synchronous linear motor using finite element simulation software; the finite element simulation software includes: ANSYS, COMSOL or FEMM;

[0050] A2: Define the material properties in the model; the material properties include: primary core, secondary plate, permanent magnet, air gap, armature core, armature winding and air medium;

[0051] A3: Define the boundary conditions and excitation sources of the model.

[0052] See also Figure 2, marking sampling points on the running track of the permanent magnet synchronous linear motor mover in the finite element model, including:

[0053] B1: extracting the running track length range of the mover of the target permanent magnet synchronous linear motor; wherein the target permanent magnet synchronous linear motor refers to the permanent magnet synchronous linear motor to be detected;

[0054] B2: Call the lhsdesign function in MATLAB to perform Latin hypercube sampling within the orbit length range to generate sampling points;

[0055] B3: Based on the sampling points, the position of each sampling point is marked in the finite element model to obtain the position point data; wherein the position point data includes: the specific position of the sampling point on the moving track and the distance between the sampling point and the initial moving position.

[0056] Training artificial intelligence models based on location point data and corresponding electromagnetic data, including:

[0057] Integrate the electromagnetic data corresponding to the sampling points in the position point data into standard input data, and integrate the position point data into standard output data;

[0058] The artificial intelligence model is trained based on standard input data and standard output data to obtain a mover position recognition model; wherein the artificial intelligence model includes: a convolutional neural network or a deep belief network.

[0059] The electromagnetic data acquired in real time is identified through the mover position identification model, including:

[0060] The electromagnetic data of the permanent magnet synchronous linear motor during operation, which is collected in real time, is input into the mover position identification model to obtain an identification result; wherein the identification result includes: the specific position of the actual mover on the mover running track in the finite element model and the distance between the actual mover and the initial mover position.

[0061] Determine whether an early warning is needed based on the identification results, including:

[0062] Extract the distance between the actual mover and the initial mover position in the recognition result and mark it as the travel distance;

[0063] Determine whether the travel distance exceeds the preset travel distance range within the same time period; if yes, generate an early warning message and send it to the client; if no, continue to detect and judge.

[0064] For example, suppose there is a permanent magnet linear synchronous motor (PMLSM) and the position of its mover needs to be monitored. The method for achieving stroke self-detection is as follows:

[0065] Step 1: Obtain electromagnetic data of the permanent magnet synchronous linear motor during operation;

[0066] 1. Data acquisition equipment:

[0067] Use data acquisition equipment to collect electromagnetic data of the permanent magnet synchronous linear motor during operation.

[0068] Install current sensor, voltage sensor, back EMF sensor and air gap magnetic flux density sensor.

[0069] Connect the sensor to the data acquisition device to ensure stable signal transmission.

[0070] The current, voltage, back electromotive force and air gap flux density data of the motor are collected when it is running and stored in the memory of the computer or controller.

[0071] 2. Electromagnetic data type:

[0072] Current: The current data in the winding is collected through the current sensor.

[0073] Voltage: The voltage data at both ends of the winding is collected through a voltage sensor.

[0074] Back EMF: The back EMF data in the winding is collected through the back EMF sensor.

[0075] Air gap magnetic density: The magnetic flux density data in the air gap is collected through an air gap magnetic density sensor (such as a Hall effect sensor).

[0076] Step 2: Construct a finite element model of the permanent magnet synchronous linear motor based on finite element simulation software;

[0077] 1. Model construction:

[0078] The 2D or 3D model of the permanent magnet synchronous linear motor is constructed by finite element simulation software such as ANSYS, and modeled according to the actual geometry and size of the motor.

[0079] 2. Define material properties:

[0080] Primary core: Use silicon steel sheets (such as DW21-50).

[0081] Secondary board: Use magnetic conductive material.

[0082] Permanent magnet: Use neodymium iron boron (such as NdFe35).

[0083] Air Gap: Use air medium.

[0084] Armature core: Use silicon steel sheets.

[0085] Armature Winding: Copper wire is used.

[0086] 3. Define boundary conditions and excitation sources:

[0087] Boundary conditions: Set periodic boundary conditions, symmetric boundary conditions, etc.

[0088] Excitation source: three-phase AC current:

[0089] Three-phase alternating current, for example:

[0090] iA(t)=Im*sin(ωt);

[0091] iB(t)=Im*sin(ωt-120°);

[0092] iC(t)=Im*sin(ωt+120°);

[0093] Where Im is the current amplitude and ω is the angular frequency.

[0094] Step 3: Mark sampling points on the running track of the permanent magnet synchronous linear motor mover in the finite element model to obtain position point data;

[0095] 1. Extract the running track length range:

[0096] Identify the permanent magnet linear synchronous motors that need to be monitored.

[0097] Extract the running track length range of the mover. For example, assume that the running track length of the mover is 100 mm.

[0098] 2. Sampling point generation:

[0099] Use the lhsdesign function in MATLAB to perform Latin hypercube sampling within the orbit length range to generate sampling points.

[0100] For example, 100 sampling points are generated, which are evenly distributed between 0 and 100 mm.

[0101] 3. Mark the sampling points:

[0102] Mark the location of each sampling point in the finite element model.

[0103] Record the specific position of each sampling point on the mover's running track and the distance between the sampling point and the initial mover position.

[0104] For example, sampling point 1 is located at 1 mm and sampling point 2 is located at 2 mm.

[0105] Step 4: Calculate the electromagnetic data corresponding to the sampling point in each position point data through the finite element simulation algorithm, and train the artificial intelligence model based on the position point data and the corresponding electromagnetic data to obtain the mover position recognition model;

[0106] 1. Simulation calculation:

[0107] Set the solver type and select the Transient solver.

[0108] Mesh the model to ensure that the mesh quality meets the simulation requirements.

[0109] Set solution parameters, such as time step, solution accuracy, etc.

[0110] Run the simulation to calculate the electromagnetic data for each sampling point, including current, voltage, back electromotive force, and air gap flux density.

[0111] 2. Data Integration:

[0112] The electromagnetic data corresponding to each sampling point are integrated into standard input data.

[0113] The position data of each sampling point is integrated into the standard output data.

[0114] For example, the electromagnetic data of sampling point 1 is current 0.5A, voltage 220V, back electromotive force 10V, air gap magnetic flux density 0.1T, and position data is 10mm.

[0115] 3. Model training:

[0116] Choose an appropriate AI model, such as a convolutional neural network (CNN) or a deep belief network (DBN).

[0117] For example, the deep belief network (DBN) model is trained using standard input data and standard output data to obtain a mover position recognition model.

[0118] The cross-validation method was used to evaluate the performance of the model and select the best model parameters.

[0119] Step 5: Identify the electromagnetic data acquired in real time through the mover position recognition model to obtain the recognition result; determine whether an early warning is needed based on the recognition result;

[0120] 1. Real-time monitoring:

[0121] The electromagnetic data of the motor during operation is collected in real time through sensors.

[0122] The real-time collected data is preprocessed by filtering, denoising, etc. to improve the data quality.

[0123] 2. Location Identification:

[0124] Extract electromagnetic data such as current, voltage, back EMF, and air gap flux density from real-time acquired data.

[0125] The extracted electromagnetic data is input into the trained mover position recognition model, and the model outputs the predicted distance of the mover from the initial position of the mover.

[0126] For example, the electromagnetic data collected in real time are current 0.55A, voltage 222V, back electromotive force 12V, air gap magnetic flux density 0.15T, and the position predicted by the model is 15mm.

[0127] 3. Position judgment and early warning:

[0128] The distance between the actual mover and the initial mover position in the recognition result is extracted and marked as the stroke distance, which is 15 mm.

[0129] A preset travel distance range is extracted and set within the same time period. For example, the preset travel distance range within the same time period is between 16 mm and 18 mm.

[0130] Since the stroke distance is 15 mm, it exceeds the preset stroke distance range set in the same time period; an early warning message is generated to indicate that the mover position is out of the normal range.

[0131] Working principle of the present invention:

[0132] The present invention obtains electromagnetic data of a permanent magnet synchronous linear motor when it is running; constructs a finite element model of the permanent magnet synchronous linear motor based on finite element simulation software; obtains position point data by marking sampling points on the running track of a mover of the permanent magnet synchronous linear motor in the finite element model; calculates electromagnetic data corresponding to the sampling points in each position point data by a finite element simulation algorithm, and trains an artificial intelligence model based on the position point data and the corresponding electromagnetic data to obtain a mover position recognition model; recognizes the electromagnetic data obtained in real time based on the mover position recognition model to obtain a recognition result; and determines whether an early warning is needed based on the recognition result, thereby solving the problem that the existing PMLSM stroke self-detection method mainly relies on position sensors and methods based on single electromagnetic parameters or mathematical models, and although these methods can provide certain position information, they have the problems of low reliability, high cost, difficulty in installation and maintenance, insufficient accuracy and robustness in complex environments; and at the same time, in practical applications, field data collection requires complex sensor arrangements and data acquisition systems, which increases the complexity and cost of the system, and the field collected data may be interfered by environmental noise, resulting in problems affecting data quality and model training effects.

[0133] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A segmented PMLSM stroke self-detection method, characterized in that: include: Step 1: Obtain electromagnetic data of the permanent magnet synchronous linear motor during operation; Step 2: Construct a finite element model of the permanent magnet synchronous linear motor based on finite element simulation software; Step 3: Mark sampling points on the running track of the permanent magnet synchronous linear motor mover in the finite element model to obtain position point data; Step 4: Calculate the electromagnetic data corresponding to the sampling point in each position point data through the finite element simulation algorithm, and train the artificial intelligence model based on the position point data and the corresponding electromagnetic data to obtain the mover position recognition model; Step 5: Identify the electromagnetic data acquired in real time through the mover position recognition model to obtain the recognition result; determine whether an early warning is needed based on the recognition result.

2. A segmented PMLSM stroke self-detection method according to claim 1, characterized in that: The step of obtaining electromagnetic data of the permanent magnet linear synchronous motor during operation includes: The electromagnetic data of the permanent magnet synchronous linear motor during operation is collected by a data acquisition device; wherein the electromagnetic data includes: current, voltage, back electromotive force and air gap flux density.

3. A segmented PMLSM stroke self-detection method according to claim 1, characterized in that: The finite element model of the permanent magnet synchronous linear motor is constructed based on the finite element simulation software, including: A1: Build a 2D or 3D model of the permanent magnet synchronous linear motor using finite element simulation software; the finite element simulation software includes: ANSYS, COMSOL or FEMM; A2: Define the material properties in the model; A3: Define the boundary conditions and excitation sources of the model.

4. A segmented PMLSM stroke self-detection method according to claim 1, characterized in that: The step of marking sampling points on the running track of the permanent magnet synchronous linear motor mover in the finite element model includes: B1: extracting the running track length range of the mover of the target permanent magnet synchronous linear motor; wherein the target permanent magnet synchronous linear motor refers to the permanent magnet synchronous linear motor to be detected; B2: Call the lhsdesign function in MATLAB to perform Latin hypercube sampling within the orbit length range to generate sampling points; B3: Based on the sampling points, the position of each sampling point is marked in the finite element model to obtain the position point data; wherein the position point data includes: the specific position of the sampling point on the moving track and the distance between the sampling point and the initial moving position.

5. A segmented PMLSM stroke self-detection method according to claim 1, characterized in that: The training of the artificial intelligence model based on the location point data and the corresponding electromagnetic data includes: Integrate the electromagnetic data corresponding to the sampling points in the position point data into standard input data, and integrate the position point data into standard output data; The artificial intelligence model is trained based on standard input data and standard output data to obtain a mover position recognition model; wherein the artificial intelligence model includes: a convolutional neural network or a deep belief network.

6. A segmented PMLSM stroke self-detection method according to claim 1, characterized in that: The identifying of electromagnetic data acquired in real time by using the mover position identification model includes: The electromagnetic data of the permanent magnet synchronous linear motor during operation collected in real time is input into the mover position identification model to obtain an identification result; wherein the identification result includes: the specific position of the actual mover on the mover running track in the finite element model and the distance of the actual mover from the initial mover position.

7. A segmented PMLSM stroke self-detection method according to claim 1, characterized in that: The determining whether an early warning is required based on the recognition result includes: Extract the distance between the actual mover and the initial mover position in the recognition result and mark it as the travel distance; Determine whether the travel distance exceeds the preset travel distance range within the same time period; if yes, generate an early warning message and send it to the client; if no, continue to detect and judge.

Images