Online detection method, device, terminal and medium for dynamic performance of linear motor

Abstract

The present invention discloses an online detection method, device, terminal and storage medium for the dynamic performance of a linear motor, which collects real physical data of the linear motor, generates simulation verification data of the linear motor through a simulation model of the linear motor, performs a preset data fusion operation on the real physical data and the simulation verification data to obtain fused data, and calculates the dynamic performance index data of the linear motor based on the fused data to complete the online detection of the dynamic performance of the linear motor. When collecting the real physical data of the linear motor, the corresponding simulation verification data is generated, and then the real physical data and the simulation verification data are fused, and finally the dynamic performance index data of the linear motor is calculated based on the fused data, so that the collected data can be corrected based on the data fusion, which is conducive to eliminating the error of the linear motor data collection, improving the accuracy of the data collection, and finally improving the accuracy of the linear motor performance detection.

Description

Technical Field

[0001] The present invention relates to the technical field of linear motors, and in particular to an online detection method, device, terminal and storage medium for dynamic performance of a linear motor. Background Art

[0002] Traditional electric motors usually move in the form of rotation when working, but most parts driven by rotating electric motors need to move in a straight line, such as electric motor-driven trams. For this reason, traditional electric motor-driven machines often need to have a transmission mechanism inside the machine that converts rotational motion into linear motion. The introduction of the transmission mechanism will inevitably affect the system's response speed, progress, dynamic strength and other performance. Decades ago, people have proposed the idea: Can a linear motion motor be used directly to drive the machine, so that the intermediate transmission mechanism can be omitted, thereby improving the performance of the machine. For this reason, a linear motor that can directly convert electrical energy into linear motion mechanical energy was born.

[0003] A linear motor can be considered as a rotating motor cut radially and unfolded into a plane. The side evolved from the stator is called the primary, and the side evolved from the rotor is called the secondary. Alternating current is passed through the primary, and the secondary moves linearly along the primary under the action of electromagnetic force. At this time, it is usually necessary to ensure that the primary is long enough to extend to the position required for movement, while the secondary does not need to be that long. In actual applications, the primary and secondary are manufactured to different lengths to ensure that the coupling between the primary and secondary remains unchanged within the required travel range. Linear motors can have a short primary and a long secondary, or a long primary and a short secondary. Considering the manufacturing cost and operating expenses, a short primary and a long secondary are generally used.

[0004] Since the linear motor eliminates the transmission mechanism of the traditional rotary motor, the linear motor can achieve the "zero transmission" transmission mode, which makes the linear motor have many performance indicators and advantages that the traditional rotary motor cannot achieve, such as high-speed response, high precision, high dynamic stiffness, high speed, etc. For this reason, linear motors are mostly used in machines with high performance requirements, such as maglev trains. The linear motor is an extremely complex system. The mutual influence between electricity, force, heat, magnetism and other factors in its working process is a very complex process. The maturity of linear motor technology still requires a lot of theoretical research and practical application work. Since the mutual influence between electricity, force, heat, magnetism and other factors in the working process of the linear motor is extremely complex, in the actual application or research of the linear motor, it is often necessary to collect data on electricity, force, heat, magnetism and many other aspects of the linear motor in order to achieve comprehensive monitoring of the working process of the linear motor and adjust the control of the linear motor in time to achieve good working performance. It can be seen that the accurate collection of the performance data of the linear motor is an important part of the linear motor technology. However, in practice, it is found that due to the complex electrical, force, thermal, and magnetic mechanisms of linear motors and the volatile data acquisition environment, the current data acquisition results for linear motors are not accurate, resulting in inaccurate test results for linear motor performance. Summary of the invention

[0005] The technical problem to be solved by the present invention is that the accuracy of the current data collection results for the linear motor is not high, resulting in inaccurate detection results of the linear motor performance.

[0006] In order to solve at least one of the above technical problems, the first aspect of the present invention discloses an online detection method for the dynamic performance of a linear motor, the method comprising:

[0007] Collect real physical data of linear motors;

[0008] Generating simulation verification data of the linear motor through a simulation model of the linear motor;

[0009] Performing a preset data fusion operation on the real physical data and the simulation verification data to obtain fused data;

[0010] Dynamic performance index data of the linear motor is calculated based on the fusion data to complete online detection of the dynamic performance of the linear motor.

[0011] As an optional implementation, in the first aspect of the present invention, the simulation model is a traditional mathematical model established based on the coupling relationship between the real physical data.

[0012] As an optional implementation, in the first aspect of the present invention, the simulation model is a neural network model established based on the coupling relationship between the real physical data.

[0013] As an optional embodiment, in the first aspect of the present invention, the method further comprises:

[0014] Count the amount of real physical data collected;

[0015] When the data volume is greater than a preset data volume threshold, a new training set is generated based on the collected real physical data, and the neural network model is retrained based on the new training set to update the neural network model.

[0016] As an optional embodiment, in the first aspect of the present invention, the method further comprises:

[0017] Calculating the deviation between the real physical data and the simulation verification data;

[0018] When the deviation is greater than a preset deviation threshold, a new training set is generated based on the collected real physical data, and the neural network model is retrained based on the new training set to update the neural network model.

[0019] As an optional implementation, in the first aspect of the present invention, generating a new training set based on the collected real physical data includes:

[0020] Calculate the Euclidean distance between the real physical data and the training data in the old training set;

[0021] The training data with the smallest Euclidean distance to the real physical data is determined as the data to be replaced;

[0022] The data to be replaced in the old training set is replaced with real physical data to generate a new training set.

[0023] As an optional implementation, in the first aspect of the present invention, performing a preset data fusion operation on the real physical data and the simulation verification data to obtain fused data includes:

[0024] Determining respective trust values ​​of the real physical data and the simulation verification data;

[0025] Calculating a normalization coefficient based on the trust values ​​of the real physical data and the simulation verification data;

[0026] The real physical data and the simulation verification data are fused based on the normalization coefficient, the trust values ​​of the real physical data and the simulation verification data to obtain fused data.

[0027] The second aspect of the present invention discloses an online detection device for the dynamic performance of a linear motor, the device comprising:

[0028] The acquisition module is used to collect the real physical data of the linear motor;

[0029] A simulation verification module, used for generating simulation verification data of the linear motor through a simulation model of the linear motor;

[0030] A data fusion module, used for performing a preset data fusion operation on the real physical data and the simulation verification data to obtain fused data;

[0031] The performance index calculation module is used to calculate the dynamic performance index data of the linear motor based on the fusion data to complete the online detection of the dynamic performance of the linear motor.

[0032] The third aspect of the present invention discloses an online detection terminal for the dynamic performance of a linear motor, the terminal comprising:

[0033] A memory storing executable program code;

[0034] a processor coupled to the memory;

[0035] The processor calls the executable program code stored in the memory to execute part or all of the steps in the online detection method of the dynamic performance of the linear motor disclosed in the first aspect of the present invention.

[0036] The fourth aspect of the present invention discloses a computer storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute some or all of the steps in the online detection method of the dynamic performance of a linear motor disclosed in the first aspect of the present invention.

[0037] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0038] In an embodiment of the present invention, real physical data of a linear motor is collected, simulation verification data of the linear motor is generated through a simulation model of the linear motor, a preset data fusion operation is performed on the real physical data and the simulation verification data to obtain fused data, and dynamic performance index data of the linear motor is calculated based on the fused data to complete online detection of the dynamic performance of the linear motor. When collecting the real physical data of the linear motor, corresponding simulation verification data is generated, and then the real physical data and the simulation verification data are fused, and finally the dynamic performance index data of the linear motor is calculated based on the fused data, so that the collected data can be corrected based on the fusion of the real data and the simulation data of the linear motor, which is conducive to eliminating errors in data collection of the linear motor, improving the accuracy of data collection, and ultimately improving the accuracy of performance detection of the linear motor. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. 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 creative work.

[0040] Figure 1 It is a flow chart of an online detection method of dynamic performance of a linear motor disclosed in an embodiment of the present invention;

[0041] Figure 2 is a comparison diagram of the simulated temperature and the measured temperature of a certain point in an embodiment of the present invention;

[0042] Figure 3 It is a structural schematic diagram of an online detection device for dynamic performance of a linear motor disclosed in an embodiment of the present invention;

[0043] Figure 4 It is a structural schematic diagram of an online detection terminal for dynamic performance of a linear motor disclosed in an embodiment of the present invention;

[0044] Figure 5 It is a structural schematic diagram of a computer storage medium disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0045] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only 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.

[0046] The terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish different objects rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, device, product or end including a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units inherent to these processes, methods, products or ends.

[0047] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0048] The present invention discloses an online detection method, device, terminal and storage medium for the dynamic performance of a linear motor, which collects real physical data of the linear motor, generates simulation verification data of the linear motor through a simulation model of the linear motor, performs a preset data fusion operation on the real physical data and the simulation verification data to obtain fused data, and calculates the dynamic performance index data of the linear motor based on the fused data to complete the online detection of the dynamic performance of the linear motor. By generating corresponding simulation verification data when collecting the real physical data of the linear motor, then fusing the real physical data and the simulation verification data, and finally calculating the dynamic performance index data of the linear motor based on the fused data, the collected data can be corrected based on the fusion of the real data and the simulation data of the linear motor, which is conducive to eliminating the error of linear motor data collection, improving the accuracy of data collection, and finally improving the accuracy of linear motor performance detection. The following are detailed descriptions.

[0049] Embodiment 1

[0050] See also Figure 1 , Figure 1FIG. 1 is a flow chart of an online detection method for the dynamic performance of a linear motor disclosed in an embodiment of the present invention. Figure 1 As shown, the online detection method of the dynamic performance of the linear motor may include the following operations:

[0051] 101. Collect real physical data of linear motor.

[0052] In the above step 101, different sensing devices can be deployed in the linear motor system to collect different types of data of the linear motor. For example, a laser interferometer can be deployed to collect displacement data of the linear motor feed system in the feed direction; an infrared thermal imager can be deployed to collect temperature data of the linear motor; an ammeter can be deployed to collect current data of the linear motor, etc. Usually, the mutual influence between electricity, force, heat, magnetism and other factors during the operation of the linear motor is extremely complex, so in order to achieve comprehensive monitoring of the linear motor, it is generally necessary to deploy a variety of sensing devices.

[0053] 102. Generate simulation verification data of the linear motor using the simulation model of the linear motor.

[0054] In the above step 102, while the deployed sensors collect real physical data, the pre-established simulation model of the linear motor can be used to generate corresponding simulation verification data. For example, while the displacement data of the feed system in the feed direction is collected by the laser interferometer, the corresponding displacement data at this time is generated by the simulation model.

[0055] 103. Perform a preset data fusion operation on the real physical data and the simulation verification data to obtain fused data.

[0056] In the above step 103, although the real physical data collected by the sensor can most directly reflect the working state of the linear motor, it is usually affected by many accidental factors, such as the surrounding environment, the working state of the sensor, etc., so that the collected real physical data often deviates from the correct data. As for the data generated by the simulation model, it is only necessary to ensure the accuracy of the simulation model modeling. The data generated by the simulation model is usually more accurate and is basically not affected by accidental factors. Figure 2 As shown in the figure, the simulated temperature at point d of the linear motor system shows obvious linearity, while the measured temperature at point d fluctuates around the simulated temperature curve due to the influence of accidental factors. Therefore, the real physical data and simulation verification data can be integrated to combine the characteristics of the two types of data, thereby reducing the accidental deviation of the real physical data and reducing the deviation of the simulation verification data due to imperfect modeling, thereby improving the accuracy of data acquisition.

[0057] 104. Calculate dynamic performance index data of the linear motor based on the fusion data to complete online detection of the dynamic performance of the linear motor.

[0058] In the above step 104, the online detection of the dynamic performance of the linear motor can be achieved by calculating the dynamic performance index data of the linear motor. For example, the dynamic performance index data of the linear motor may include feed acceleration, positioning accuracy, temperature, response speed, feed speed regulation range, feed speed, transmission stiffness, etc. By using the mathematical relationship between the fused data and the dynamic performance index data, the dynamic performance index data can be calculated based on the fused data. For example, the feed acceleration is related to electromagnetic reasoning, the friction of the guide rail, and the total mass of the moving parts, that is, the feed acceleration can be calculated based on the fused data related to electromagnetic reasoning, the friction of the guide rail, and the total mass of the moving parts.

[0059] In an optional embodiment, the simulation model is a traditional mathematical model established based on the coupling relationship between the real physical data.

[0060] In this optional embodiment, although the mutual influence between electricity, force, heat, magnetism and other factors during the operation of the linear motor is a very complex process, a fixed mathematical model can still be used to represent the coupling relationship between electricity, force, heat, magnetism and other factors. In the traditional mathematical model, the coupling relationship between the transmission stiffness and other factors can be expressed as:

[0061] K=4π 2 Mf 2

[0062] The coupling relationship between damping and other factors can be expressed as:

[0063] c=4π 2 Mfξ

[0064] Among them, K represents transmission stiffness, c represents damping, M represents system mass, f represents natural frequency, and ξ represents damping ratio.

[0065] The coupling relationship between the system dynamic stiffness and other factors can be expressed as:

[0066]

[0067]

[0068] Among them, K p is the position loop gain, K v is the speed loop gain, T n is the speed loop integral response time constant, m is the mass of the worktable, workpiece and linear motor moving parts, K f is the force constant.

[0069] It can be seen that in the traditional mathematical model, the coupling relationship between various factors of the linear motor will be expressed as a mathematical relationship. By integrating these mathematical relationships, the simulation model of the linear motor can be obtained, thereby simulating the working process of the linear motor. Specifically, when establishing the traditional mathematical model of the linear motor, the mathematical model of the linear motor can be established from different levels such as the mechanical structure, electrical relationship, and control model of the linear motor, so as to express all the coupling relationships of the linear motor as accurately as possible.

[0070] In an optional embodiment, the simulation model is a neural network model established based on the coupling relationship between the real physical data.

[0071] In this optional embodiment, the neural network model is an algorithm with a clear structure, simple algorithm and nonlinear adaptive data processing capability compared to other algorithms. It has strong generalization ability and nonlinear mapping ability, fast data transmission speed, and parallelism from the perspective of functional simulation. Combined with the complex and changeable characteristics of the linear motor system, the neural network model has advantages in predicting its highly nonlinear dynamic performance changes. Specifically, when it is necessary to use the neural network model to express the coupling relationship between real physical data, the two real physical data can be used as the input and output of the neural network model respectively, and then the neural network model is trained using a sufficient amount of the two real physical data as a training set. When the neural network model reaches a convergence state, the coupling relationship between the two real physical data can be expressed. Relatively speaking, the neural network model and the traditional mathematical model each have their advantages and disadvantages. The dynamic performance of the linear motor system during operation is usually changing. For example, the running-in, adjustment, aging, fatigue and other factors between the components during the operation of the linear motor will cause the dynamic performance of the linear motor to change. The traditional mathematical model is usually static, which leads to its inability to well reflect the changes in the dynamic performance of the linear motor system during operation, which is not enough to meet the current high standards and high requirements. Although the neural network model can adapt to the changes in the dynamic performance of the linear motor system during operation, it usually requires a large number of high-quality training sets for training to achieve a relatively ideal state. To this end, the advantages and disadvantages of the two models can be combined. Specifically, the neural network model and the traditional mathematical model can be used at the same time to generate simulation verification data, and finally the simulation data generated by the two models are used as the source data for data fusion. Alternatively, only the traditional mathematical model can be used for simulation in the early stage of the linear motor operation. When the amount of collected data is large enough, the traditional mathematical model can be replaced by the neural network model. In this way, the advantages of the two models can be combined, thereby improving the quality of the final fusion data.

[0072] In an optional embodiment, the method further includes:

[0073] Count the amount of real physical data collected;

[0074] When the data volume is greater than a preset data volume threshold, a new training set is generated based on the collected real physical data, and the neural network model is retrained based on the new training set to update the neural network model.

[0075] In this optional embodiment, as mentioned above, the dynamic performance of the linear motor system usually changes with the change of the system running time. Therefore, in order to enable the neural network model to adapt to the change of the dynamic performance of the linear motor system, the neural network model needs to be retrained to update the neural network model. Specifically, a time-based model update method can be used to measure the running time of the linear motor using the amount of data of the real physical data that has been collected. When the amount of data is greater than the data amount threshold, it means that the linear motor system has been running for a long enough time since the last update of the neural network model, and the dynamic performance of the linear motor system may have changed. At this time, updating the neural network model can better adapt to the changes of the linear motor system.

[0076] It can be seen that by implementing this optional embodiment, a time-based model updating method is adopted, and the amount of real physical data that has been collected is used to measure the running time of the linear motor, so as to achieve the effect of updating the neural network model based on the running time of the linear motor, so that the neural network model can adapt to the changes in the dynamic performance of the linear motor system, improve the accuracy of the simulation data, and then improve the accuracy of the fusion data.

[0077] In an optional embodiment, the method further includes:

[0078] Calculating the deviation between the real physical data and the simulation verification data;

[0079] When the deviation is greater than a preset deviation threshold, a new training set is generated based on the collected real physical data, and the neural network model is retrained based on the new training set to update the neural network model.

[0080] In this optional embodiment, a model updating method based on prediction error is also proposed. The deviation between the simulation data output by the neural network model and the real physical data is calculated. When the deviation is greater than the preset deviation threshold, it indicates that the current neural network model can no longer adapt to the changes in the dynamic performance of the current linear motor system. At this time, the neural network model needs to be updated to better adapt to the changes in the linear motor system.

[0081] It can be seen that by implementing this optional embodiment and adopting a model updating method based on prediction error, when the deviation between the simulation data and the real physical data is greater than the preset deviation threshold, the neural network model is updated, so that the neural network model can adapt to changes in the dynamic performance of the linear motor system, improve the accuracy of the simulation data, and further improve the accuracy of the fusion data.

[0082] In an optional embodiment, generating a new training set based on the collected real physical data includes:

[0083] Calculate the Euclidean distance between the real physical data and the training data in the old training set;

[0084] The training data with the smallest Euclidean distance to the real physical data is determined as the data to be replaced;

[0085] The data to be replaced in the old training set is replaced with real physical data to generate a new training set.

[0086] In this optional embodiment, when the neural network model needs to be retrained, it is necessary to use a training set containing new features, that is, it is necessary to include the features of the collected real physical data into the training set. However, in order to ensure the training speed of the neural network model, the size of the training set needs to be kept unchanged. Therefore, it is necessary to delete the data samples in the original training set and replace them with data containing new features. When the original training data samples closest to the new feature data are deleted, the impact on the feature description of the neural network model is minimized. Therefore, the method of calculating the Euclidean distance can be used to calculate the distance between the new test data sample (i.e., the real physical data) and each data training sample in the training set (i.e., the training data in the old training set), and the training data sample closest to the new feature is closest to the test data sample. The training data sample closest to the test data sample is selected for data replacement operation, so that the update of the neural network model with less feature description impact and faster training speed can be achieved.

[0087] In an optional embodiment, performing a preset data fusion operation on the real physical data and the simulation verification data to obtain fused data includes:

[0088] Determining respective trust values ​​of the real physical data and the simulation verification data;

[0089] Calculating a normalization coefficient based on the trust values ​​of the real physical data and the simulation verification data;

[0090] The real physical data and the simulation verification data are fused based on the normalization coefficient, the trust values ​​of the real physical data and the simulation verification data to obtain fused data.

[0091] In this optional embodiment, the trust values ​​of the real physical data and the simulation verification data can be given by their respective trust value functions. The trust value function can be used to assign a trust value to each real physical data or simulation verification data. The trust value is between 0 and 1. The sum of the trust values ​​of all real physical data is 1, and the sum of the trust values ​​of all simulation verification data is 1. The role of the trust value function is to assign trust to each real physical data or simulation verification data. For example, for a certain real physical data, if the trust value assigned by the trust value function is 0.9, it means that the credibility of the real physical data is high. If the trust value assigned by the trust value function is 0.1, it means that the credibility of the real physical data is low. The trust value function can be set by engineers according to actual conditions to achieve better results.

[0092] The normalization coefficient is calculated by the following formula (1):

[0093]

[0094] Among them, K is the normalization coefficient, E is the real physical data, F is the simulation verification data, m1(E) is the trust value of the real physical data, and m2(F) is the trust value of the simulation verification data.

[0095] The real physical data and simulation verification data are fused using the following formula (2):

[0096]

[0097] Among them, K is the normalization coefficient, A is the fusion data, E is the real physical data, F is the simulation verification data, m(A) is the trust value of the fusion data, m1(E) is the trust value of the real physical data, and m2(F) is the trust value of the simulation verification data.

[0098] The data fusion process in the following table is illustrated as an example:

[0099]

[0100] As shown in the table above, for a certain performance data of the linear motor (such as temperature), the collected real physical data show that the temperature of 21.8 is likely to be 0.2, 21.9 is likely to be 0.4, and 22.1 is likely to be 0.4. The collected simulation verification data show that the temperature of 21.8 is likely to be 0.4, 21.9 is likely to be 0.2, 22.0 is likely to be 0.2, and 22.2 is likely to be 0.2.

[0101] The normalization coefficient is calculated by formula 1:

[0102]

[0103] Then, the real physical data and simulation verification data are fused by formula 2 to obtain the possibility that the fused data considers the temperature to be 21.8:

[0104]

[0105] Similarly, the fused data gives the following possibilities for the temperature to be 21.9, 22.0, 22.1, and 22.2:

[0106] m(A=21.9)=0.5

[0107] m(A=22.0)=0

[0108] m(A=22.1)=0

[0109] m(A=22.2)=0

[0110] Finally, the mathematical expectation of the possibility of fused data distribution is taken as the result of data fusion, that is, 21.8*0.5+21.9*0.5=21.85, so the final temperature after data fusion is 21.85.

[0111] It can be seen that implementing this optional embodiment, assigning trust values ​​to real physical data and simulation verification data, and using the trust values ​​to fuse the real physical data and simulation verification data through a preset data fusion formula is conducive to ensuring the accuracy of data fusion.

[0112] It can be seen that the implementation Figure 1 The described online detection method of the dynamic performance of a linear motor collects real physical data of the linear motor, generates simulation verification data of the linear motor through a simulation model of the linear motor, performs a preset data fusion operation on the real physical data and the simulation verification data to obtain fused data, and calculates the dynamic performance index data of the linear motor based on the fused data to complete the online detection of the dynamic performance of the linear motor. By generating corresponding simulation verification data when collecting the real physical data of the linear motor, then fusing the real physical data and the simulation verification data, and finally calculating the dynamic performance index data of the linear motor based on the fused data, the collected data can be corrected based on the fusion of the real data and the simulation data of the linear motor, which is conducive to eliminating the error of linear motor data collection, improving the accuracy of data collection, and ultimately improving the accuracy of linear motor performance detection.

[0113] Embodiment 2

[0114] See also Figure 3 , Figure 3Schematic diagram of the structure of an online detection device for the dynamic performance of a linear motor disclosed in an embodiment of the present invention. Figure 3 As shown, the online detection device for the dynamic performance of the linear motor may include:

[0115] The acquisition module 301 is used to acquire the real physical data of the linear motor;

[0116] A simulation verification module 302, used to generate simulation verification data of the linear motor through a simulation model of the linear motor;

[0117] The data fusion module 303 is used to perform a preset data fusion operation on the real physical data and the simulation verification data to obtain fused data;

[0118] The performance index calculation module 304 is used to calculate the dynamic performance index data of the linear motor based on the fusion data to complete the online detection of the dynamic performance of the linear motor.

[0119] For the specific description of the online detection device for the dynamic performance of the linear motor, reference may be made to the specific description of the online detection method for the dynamic performance of the linear motor, which will not be described in detail here.

[0120] Embodiment 3

[0121] See also Figure 4 , Figure 4 Schematic diagram of the structure of an online detection terminal for the dynamic performance of a linear motor disclosed in an embodiment of the present invention. Figure 4 As shown, the online detection terminal of the dynamic performance of the linear motor may include:

[0122] A memory 401 storing executable program codes;

[0123] a processor 402 coupled to the memory 401;

[0124] The processor 402 calls the executable program code stored in the memory 401 to execute the steps of the online detection method of the dynamic performance of the linear motor disclosed in the first embodiment of the present invention.

[0125] Embodiment 4

[0126] An embodiment of the present invention discloses a computer storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute the steps in the online detection method of the dynamic performance of a linear motor disclosed in the first embodiment of the present invention.

[0127] The device embodiments described above are only illustrative, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, i.e., they may be located in one place, or they may be distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative work.

[0128] Through the specific description of the above embodiments, 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, it can also be implemented by hardware. Based on such an understanding, the above technical solution can be essentially or partly contributed to the prior art in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0129] Finally, it should be noted that the online detection method, device, terminal and storage medium for the dynamic performance of a linear motor disclosed in the embodiments of the present invention only disclose the preferred embodiments of the present invention, which are only used to illustrate the technical solution of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments can still be modified, or some of the technical features therein can be replaced by equivalents. 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. An online detection method for the dynamic performance of a linear motor, characterized in that: The method comprises: Collect real physical data of linear motors; Generate simulation verification data of the linear motor through a simulation model of the linear motor, wherein the simulation model is a neural network model established based on a coupling relationship between the real physical data; Performing a preset data fusion operation on the real physical data and the simulation verification data to obtain fused data; Calculating the dynamic performance index data of the linear motor based on the fusion data to complete the online detection of the dynamic performance of the linear motor; Count the amount of real physical data collected; When the data volume is greater than a preset data volume threshold, generating a new training set based on the collected real physical data, and retraining the neural network model based on the new training set to update the neural network model; The generating of a new training set based on the collected real physical data includes: Calculate the Euclidean distance between the real physical data and the training data in the old training set; The training data with the smallest Euclidean distance to the real physical data is determined as the data to be replaced; The data to be replaced in the old training set is replaced with real physical data to generate a new training set.

2. The online detection method of the dynamic performance of a linear motor according to claim 1, characterized in that: The method further comprises: Calculating the deviation between the real physical data and the simulation verification data; When the deviation is greater than a preset deviation threshold, a new training set is generated based on the collected real physical data, and the neural network model is retrained based on the new training set to update the neural network model.

3. The online detection method for the dynamic performance of a linear motor according to claim 1 or 2, characterized in that: The performing a preset data fusion operation on the real physical data and the simulation verification data to obtain fused data includes: Determining respective trust values ​​of the real physical data and the simulation verification data; Calculating a normalization coefficient based on the trust values ​​of the real physical data and the simulation verification data; The real physical data and the simulation verification data are fused based on the normalization coefficient, the trust values ​​of the real physical data and the simulation verification data to obtain fused data.

4. An online detection device for the dynamic performance of a linear motor, characterized in that: The device comprises: The acquisition module is used to collect the real physical data of the linear motor; A simulation verification module, used to generate simulation verification data of the linear motor through a simulation model of the linear motor, wherein the simulation model is a neural network model established based on a coupling relationship between the real physical data; A data fusion module, used for performing a preset data fusion operation on the real physical data and the simulation verification data to obtain fused data; A performance index calculation module, used for calculating the dynamic performance index data of the linear motor based on the fusion data, so as to complete the online detection of the dynamic performance of the linear motor; The statistical module is used to count the amount of real physical data that has been collected; An updating module, configured to generate a new training set based on the collected real physical data when the data volume is greater than a preset data volume threshold, and retrain the neural network model based on the new training set to update the neural network model; The generating of a new training set based on the collected real physical data includes: Calculate the Euclidean distance between the real physical data and the training data in the old training set; The training data with the smallest Euclidean distance to the real physical data is determined as the data to be replaced; The data to be replaced in the old training set is replaced with real physical data to generate a new training set.

5. An online detection terminal for the dynamic performance of a linear motor, characterized in that: The terminal comprises: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the online detection method for the dynamic performance of the linear motor as described in any one of claims 1-3.

6. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the online detection method for the dynamic performance of a linear motor as described in any one of claims 1 to 3 is implemented.

Images