Two-way current and voltage transmitting output method of dynamic torque sensor

Abstract

The invention discloses a two-way current and voltage transmitting output method of a dynamic torque sensor, relates to the technical field of dynamic torque sensors, and aims to solve the problem that a deviation exists between a current and voltage signal transmitted and output in the prior art and an actual torque value. The method comprises the following steps; the method comprises the following steps: S1, data acquisition and preprocessing: performing multiple experiments under different working conditions through a dynamic torque sensor, acquiring torque values and corresponding current and voltage output values as sample data, and preprocessing the sample data, including abnormal value removal and data normalization operation; and S2, constructing and training a double-path transmitting model, constructing the double-path transmitting model based on the preprocessed sample data, establishing a mapping relation between a torque value and a double-path current and voltage output value by the double-path transmitting model by adopting a multiple linear regression algorithm, and dividing the preprocessed sample data into a training set and a test set. The method has the advantage that the current and voltage transmitting output precision is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of dynamic torque sensors, more particularly to a dual-current voltage transmission output method for dynamic torque sensors. BACKGROUND

[0002] As a key device capable of measuring the torque of a rotating shaft in real time, dynamic torque sensors play an indispensable role in many fields such as industrial automation, mechanical manufacturing, automotive engineering, aerospace, etc.

[0003] The core function of a dynamic torque sensor is to convert the physical quantity of dynamic torque of a rotating shaft into measurable and transmissible electrical signals (such as current signals or voltage signals), thereby providing important data support for the state monitoring, control adjustment, and performance evaluation of the device.

[0004] In the prior art, the current-voltage transmission output method for dynamic torque sensors usually adopts a conventional linear transmission algorithm. This kind of algorithm is based on the assumption that the torque value and the current-voltage output are linearly related. By presetting fixed proportional coefficients and offsets, the collected torque signals are directly converted into corresponding current or voltage signals. However, in actual application scenarios, the working environment of dynamic torque sensors is complex, and the conventional linear transmission algorithm cannot accurately fit the nonlinear relationship between the torque value and the current-voltage output. As a result, in the case of complex torque changes, the current-voltage signals transmitted by the output deviate from the actual torque values, making it difficult to meet the demand for high-precision measurement. In view of this, we propose a dual-current voltage transmission output method for dynamic torque sensors. SUMMARY

[0005] The purpose of the present application is to provide a dual-current voltage transmission output method for dynamic torque sensors, aiming to solve the problem of deviation between the current-voltage signals transmitted by the output and the actual torque values in the prior art.

[0006] To solve the above technical problems, the present application provides the following technical scheme: a dual-current voltage transmission output method for dynamic torque sensors, which comprises the following steps:

[0007] S1, data acquisition and preprocessing, through multiple experiments under different working conditions by a dynamic torque sensor, torque values and corresponding current-voltage output values are collected as sample data, and the sample data is preprocessed, including removing outliers and data normalization operation;

[0008] S2, a two-way transmission model is constructed and trained, and a two-way transmission model is constructed based on the preprocessed sample data. The two-way transmission model uses a multiple linear regression algorithm to establish a mapping relationship between the torque value and the two-way current voltage output value. The preprocessed sample data is divided into a training set and a test set. The training set is used to train the two-way transmission model, so that the error between the current voltage value output by the two-way transmission model and the actual value of the training set gradually decreases. During the training process, after each round of training, the test set is used to evaluate the output accuracy of the two-way transmission model, and an optimized two-way transmission model is obtained.

[0009] S3, two-way current voltage transmission output, then the real-time torque signal is collected through the dynamic torque sensor and input into the optimized two-way transmission model. The two-way transmission model processes the real-time torque signal to generate two corresponding current voltage output signals respectively, realizing the transmission output of two-way current voltage.

[0010] Preferably, in the above step S1, when removing outliers, the data exceeding the range of mean value ± 3 times standard deviation is determined as outliers and removed by calculating the mean value and standard deviation of the sample data, so as to reduce the interference of abnormal data on subsequent two-way transmission model training.

[0011] Preferably, in the above step S1, when normalizing the data, the data is mapped to the interval [0, 1], and the normalization formula used is , wherein, is the original data, is the minimum value in the original data, is the maximum value in the original data, is the normalized data.

[0012] Preferably, in the above step S1, when the preprocessed sample data is divided into a training set and a test set, the training set and the test set are divided in a ratio of 80% and 20%, and the division process adopts a random sampling method.

[0013] Preferably, in the above step S3, when the real-time collected torque signal is input into the optimized two-way transmission model, the torque signal is processed in the same way as the sample data normalization operation, and when the two corresponding current voltage output signals are generated, the reverse normalization processing is performed again, and the formula is used to restore the normalized data to the actual current voltage value.

[0014] Preferably, in the above step S2, when the multiple linear regression algorithm is used to establish the mapping relationship between the torque value and the two-way current voltage output value, a regression formula is established for each of the two-way current voltage output;

[0015] The current voltage output formula of the first way is In the formula, This is the torque value. The coefficients of the regression equation for this path are... The order of the regression equation;

[0016] The current and voltage output formulas for the second channel are as follows: In the formula, This is the torque value. The coefficients of the regression equation for this path are... Let be the order of the regression equation.

[0017] Preferably, the coefficients in the current and voltage output formulas for the first path are solved using the least squares method, that is, the sum of squared errors is achieved through the least squares method. Minimum, where, The number of samples in the training set. To train the actual output values ​​of the first channel current and voltage, For the corresponding torque value, the coefficient is then determined. ;

[0018] The coefficients in the current and voltage output formulas for the second channel The same method was used to determine this.

[0019] Preferably, in step S2 above, when evaluating the output accuracy of the dual-channel transmitter model using a test set, the mean square error is used as the evaluation index. For the first channel current and voltage output, the mean square error is calculated using the following formula: In the formula, The number of samples in the test set. To test the actual output values ​​of the first channel current and voltage, This represents the predicted output values ​​of the first channel current and voltage for the dual-channel transmitter model.

[0020] For the second current and voltage output, the mean square error calculation formula is as follows: In the formula, To test the actual output values ​​of the second channel current and voltage, This is the predicted output value of the second current and voltage for the dual-channel transmitter model;

[0021] During 5 consecutive rounds of training and Training can be stopped when the decrease is less than 0.001.

[0022] Compared with the prior art, the beneficial effects of the present invention are:

[0023] 1. This invention ensures comprehensive data coverage by collecting sample data under multiple operating conditions. Combined with outlier removal and normalization processing, it provides a high-quality foundation for training the dual-channel transmitter model. At the same time, it uses a multiple linear regression algorithm to establish a mapping relationship between the dual outputs and uses the least squares method to solve the coefficients to fit the sample data to the greatest extent. This effectively solves the problem of deviation between the current and voltage signals of the transmitter output and the actual torque value in the prior art, and realizes the accurate conversion of torque signal to electrical signal.

[0024] 2. This invention adopts a dual-channel transmission model design, establishing independent regression formulas for the two current and voltage outputs respectively, supporting differentiated and accurate transmission. At the same time, in real-time signal processing, through normalization and denormalization operations consistent with the sample data, it ensures that the scale of the real-time signal is consistent with that of the training sample, and can restore the output of the dual-channel transmission model to the actual physical quantity, thereby improving the flexibility and practical application value of signal transmission.

[0025] 3. This invention accelerates the calculation of the dual-channel transmitter model through data preprocessing, and combined with a clear training stopping condition, avoids overtraining while ensuring the output accuracy of the dual-channel transmitter model, reduces unnecessary consumption of computing resources, achieves a balance between accuracy and training efficiency, and reduces the time and cost investment in practical applications. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0027] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0028] Example 1

[0029] like Figure 1 As shown, a dual-channel current-voltage transmission output method for a dynamic torque sensor includes the following steps;

[0030] S1. Data Acquisition and Preprocessing: Multiple experiments were conducted using a dynamic torque sensor under different operating conditions to collect torque values ​​and corresponding current and voltage output values ​​as sample data. The sample data was preprocessed, including outlier removal and data normalization. This ensured comprehensive data coverage by collecting samples under multiple operating conditions, reduced interference by removing outliers, and standardized the data scale by normalization. This provided a high-quality and standardized data foundation for subsequent dual-channel transmitter model training, thereby improving the reliability of the dual-channel transmitter model.

[0031] S2, a two-way transmission model is constructed and trained based on the preprocessed sample data, a two-way transmission model is constructed, a mapping relationship between the torque value and the two-way current voltage output value is established by using a multiple linear regression algorithm, the preprocessed sample data is divided into a training set and a test set, the two-way transmission model is trained by using the training set, the error between the current voltage value output by the two-way transmission model and the actual value of the training set is gradually reduced, during the training process, the output precision of the two-way transmission model is evaluated by using the test set after each round of training is completed, an optimized two-way transmission model is obtained, the mapping relationship between the torque and the current voltage is accurately fitted by using the multiple linear regression algorithm, the two-way two-way transmission model design meets the demand of simultaneously outputting two-way signals, the separation of the training set and the test set ensures the generalization ability of the two-way transmission model, iterative evaluation and optimization make the output precision of the two-way transmission model higher, and the transmission demand of different torque signals is adapted;

[0032] S3, two-way current voltage transmission output, then the real-time torque signal is collected by the dynamic torque sensor and input into the optimized two-way transmission model, the real-time torque signal is processed by the two-way transmission model, and two corresponding current voltage output signals are generated respectively, two-way current voltage transmission output is realized, real-time torque signal is quickly output as two-way current voltage after being processed by the two-way transmission model, accurate conversion of the torque signal to the electric signal is realized, and the flexibility and reliability of signal transmission are improved.

[0033] Preferably, in the above step S1, when removing the abnormal value, the data exceeding the range of the mean value ± 3 times the standard deviation is determined as the abnormal value and is removed by calculating the mean value and the standard deviation of the sample data, so as to reduce the interference of abnormal data on the subsequent two-way transmission model training, realize the identification of abnormal value, avoid the interference of abnormal value on the training of two-way transmission model, and ensure the stability of sample data, and provide a more reliable basis for two-way transmission model fitting.

[0034] Preferably, in the above step S1, when normalizing the data, the data is mapped to the [0, 1] interval, and the normalization formula used is , wherein, is the original data, is the minimum value in the original data, is the maximum value in the original data, is the normalized data, so as to eliminate the dimensional difference of different dimension data, make the torque value and the current voltage value participate in the two-way transmission model training under the same scale, speed up the calculation convergence speed, and improve the fitting precision of the two-way transmission model.

[0035] Preferably, in the above step S1, when the pre-processed sample data is divided into a training set and a test set, the ratio of 80% and 20% is divided into a training set and a test set, and the division process adopts a random sampling method, realizes that a large proportion of the training set guarantees that the double-path transmission model has enough data to learn the mapping rule, and the test set is randomly sampled to objectively evaluate the generalization ability of the double-path transmission model, avoid overfitting, and ensure the stability of the output of the double-path transmission model.

[0036] Preferably, in the above step S3, when the real-time collected torque signal is input to the optimized double-path transmission model, the torque signal is processed in the same way as the sample data normalization operation, and when the two-path corresponding current voltage output signals are generated, the reverse normalization processing is adopted, and the formula The normalized data is restored to the actual current voltage value, realizes to keep the consistency of data processing, ensures the scale of real-time signal and training sample, and the reverse normalization restores the output of the double-path transmission model to the actual physical quantity, guarantees the accuracy and practicability of the current voltage value.

[0037] Preferably, in the above step S2, when the multiple linear regression algorithm is used to establish the mapping relationship between the torque value and the double-path current voltage output value, the regression formula is established for the two-path current voltage output respectively;

[0038] The current voltage output formula of the first path is , wherein, is the torque value, is the coefficient of the regression equation of the path, is the order of the regression equation;

[0039] The current voltage output formula of the second path is , wherein, is the torque value, is the coefficient of the regression equation of the path, is the order of the regression equation, realizes the differentiated accurate transmission of the double-path signal.

[0040] Preferably, the coefficient in the current voltage output formula of the first path is solved by the least square method, that is, the least square method is used to make the error square sum minimum, wherein, is the number of training set samples, is the actual output value of the first path current voltage in the training set, is the corresponding torque value, and the coefficient is determined.

[0041] The coefficient in the current voltage output formula of the second path is The same method is used to determine, to ensure that the regression formula can be the greatest degree of sample data, to improve the double path transmission model of the torque signal transmission accuracy.

[0042] Preferably, in the above step S2, when evaluating the output accuracy of the double-path transmission model using the test set, the mean square error is used as the evaluation index, and for the first current-voltage output, the mean square error calculation formula is , wherein, is the number of test set samples, is the actual output value of the first current-voltage in the test set, is the predicted output value of the first current-voltage by the double-path transmission model;

[0043] For the second current-voltage output, the mean square error calculation formula is , wherein, is the actual output value of the second current-voltage in the test set, is the predicted output value of the second current-voltage by the double-path transmission model;

[0044] When the descending amplitude of and is less than 0.001 in five consecutive training rounds, the training is stopped, the mean square error is used to intuitively reflect the deviation between the output of the double-path transmission model and the actual value, the stop condition is set to avoid overtraining, the training efficiency is improved while ensuring the accuracy, and the output of the double-path transmission model is ensured in the optimal state of stable error.

[0045] The embodiments of the present application are disclosed, but are not limited thereto, and those skilled in the art can easily understand the spirit of the present application according to the above embodiments, and make different inferences and changes, as long as they do not deviate from the spirit of the present application, they are within the protection scope of the present application.

Claims

1. A two-path current-voltage transmission output method of a dynamic torque sensor, characterized by, The method Comprise the following steps: S1, data acquisition and pretreatment, through dynamic torque sensor in different working conditions for many times experiment, collect torque value and corresponding current voltage output value as sample data, pretreatment of sample data, including removing outliers, data normalization operation; S2, two-way transmission model construction and training, based on the pretreated sample data to build two-way transmission model, the two-way transmission model adopts multiple linear regression algorithm to establish the mapping relationship between torque value and two-way current voltage output value, the pretreated sample data is divided into training set and test set, the training set is used to train the two-way transmission model, the error between the current voltage value output by the two-way transmission model and the actual value of the training set is gradually reduced, during the training process, after completing a round of training, the test set is used to evaluate the output precision of the two-way transmission model, and the optimized two-way transmission model is obtained; S3, two-way current voltage transmission output, then through dynamic torque sensor to collect real-time torque signal, and input it to the optimized two-way transmission model, the real-time torque signal is processed by the two-way transmission model, and two corresponding current voltage output signals are generated respectively, realizing the transmission output of two-way current voltage.

2. The two-way current voltage transmitting output method of a dynamic torque sensor according to claim 1, characterized in that, In the above step S1, when removing outliers, the mean and standard deviation of sample data are calculated, the data beyond the range of mean ± 3 times standard deviation is judged as outliers and removed, so as to reduce the interference of abnormal data on subsequent two-way transmission model training.

3. The dual current voltage transmitting output method of a dynamic torque sensor according to claim 1, wherein, In the step S1, when normalizing the data, the data is mapped to the interval [0, 1] by using the normalization formula , wherein, is the original data, is the minimum value in the original data, is the maximum value in the original data, is the normalized data.

4. The dual current voltage transmitting output method of a dynamic torque sensor according to claim 1, wherein, In the above step S1, when the pretreated sample data is divided into training set and test set, it is divided into training set and test set according to the proportion of 80% and 20%, and the division process adopts random sampling method.

5. The method of claim 1, wherein the dynamic torque sensor is a two-channel current-voltage transmitter. In the step S3, the same method as the sample data normalization operation is performed on the torque signal when the real-time collected torque signal is input to the optimized two-path transmission model, and the corresponding current-voltage output signals of the two paths are generated. Then, the inverse normalization processing is performed, and the formula The normalized data is restored to the actual current-voltage value.

6. The two-way current voltage transmitting output method of a dynamic torque sensor according to claim 1, wherein, In the above step S2, when the multiple linear regression algorithm is used to establish the mapping relationship between torque value and two-way current voltage output value, the regression formula is established for two-way current voltage output respectively; The current-voltage output formula of the first path is , wherein, is a torque value, is a coefficient of the regression equation of the path, is an order of the regression equation; The current-voltage output formula of the second path is , wherein, is a torque value, is a coefficient of the regression equation of the path, is an order of the regression equation.

7. The two-way current voltage transmitting output method of a dynamic torque sensor according to claim 6, characterized in that, The coefficients in the current-voltage output formula of the first path are solved by using the least square method, that is, the sum of error squares is minimized by using the least square method , wherein is the number of training set samples, is the actual output value of the current-voltage of the first path in the training set, is the corresponding torque value, and the coefficients are determined. Coefficient in current-voltage output formula of second path Determined in the same way.

8. The two-way current voltage transmitting output method of a dynamic torque sensor according to claim 1, wherein, In the step S2, when evaluating the output accuracy of the two-way transmission model using the test set, the mean square error is used as the evaluation index. For the first current-voltage output, the mean square error calculation formula is , wherein is the number of test set samples, is the actual output value of the first current-voltage in the test set, is the predicted output value of the first current-voltage by the two-way transmission model. For the second current-voltage output, the mean square error calculation formula is , wherein, is the actual output value of the second current-voltage in the test set, is the predicted output value of the second current-voltage by the dual-channel transmission model. When the descending amplitude of the accuracy is less than 0.001 in five continuous training rounds, the training can be stopped. and the training can be stopped.

Images