A method for infusion pump motor fault detection based on an adaptive weight SVM model

Abstract

The application discloses a method for infusion pump motor fault detection based on an adaptive weight SVM model, and relates to the field of intelligent motor fault diagnosis. The method comprises the following steps: obtaining a sample set; calculating initial weights of different types of faults based on fault frequency and fault severity score; constructing a motor fault detection SVM model, and defining an objective function based on the initial weights; training the motor fault detection SVM model by using the sample set; updating the weight of the corresponding type of fault based on the misclassification loss; optimizing the objective function based on the updated weight, constructing a Lagrange function and solving it; and inputting the operation characteristics of the infusion pump motor to be detected into the motor fault detection SVM model, and outputting a detection result. The application sets different initial weights for different types of faults, and can improve the classification accuracy of the model for key faults. The application adopts an adaptive weight mechanism, adjusts the weight according to the misclassification loss, and improves the generalization ability of the model.

Description

Technical Field

[0001] This invention relates to the field of intelligent diagnosis of motor faults, specifically to a method for detecting faults in infusion pump motors based on an adaptive weighted SVM model. Background Technology

[0002] An infusion pump is an instrument that accurately controls the number of infusion drops or the infusion flow rate, ensuring that medication is delivered to the patient at a uniform rate and in an accurate and safe manner. It is commonly used in situations requiring strict control of infusion volume and medication dosage, such as when administering vasopressors, antiarrhythmic drugs, or during intravenous infusion or anesthesia in infants and young children. One of the core components of an infusion pump is the stepper motor that drives the pump system. The stepper motor precisely controls the infusion rate and volume by controlling its rotation speed and number of rotations. When the infusion pump motor malfunctions, it must be detected and addressed immediately. However, existing fault detection methods mainly rely on threshold judgment or simple statistical analysis, which have limitations when detecting complex faults. Existing technology CN111474476A discloses a motor fault prediction method, which mainly relies on an SVM model for motor fault diagnosis. However, this method does not consider the imbalance problem in the sample dataset of motor faults, i.e., the number of samples for some types of faults is much larger than that for other types of faults, and it applies the same penalty to all fault types. This can lead to classifiers potentially misclassifying test samples as the more common fault type, as the model is more influenced by the majority class during training and ignores the features of the minority class. This is especially problematic when dealing with faults that occur infrequently (few samples) but are highly dangerous, making accurate identification difficult. Summary of the Invention

[0003] The purpose of this invention is to improve the accuracy of the motor fault detection SVM model for detecting faults in infusion pump motors.

[0004] To achieve the above objectives, this invention provides a method for fault detection of infusion pump motors based on an adaptive weighted SVM model, the method comprising the following steps:

[0005] Step 1: Obtain historical data on the operating characteristics of the infusion pump motor and the corresponding operating status of the infusion pump motor to obtain a sample set;

[0006] Step 2: Calculate the initial weights for different types of faults based on fault frequency and fault severity scores;

[0007] Step 3: Construct a motor fault detection SVM model and define an objective function based on the initial weights;

[0008] Step 4: Train the motor fault detection SVM model using the sample set and calculate the misclassification loss of fault detection;

[0009] Step 5: Update the weights of the corresponding fault types based on the misclassification loss;

[0010] Step 6: Optimize the objective function based on the updated weights, construct the Lagrangian function and solve for it to obtain the classification hyperplane parameters;

[0011] Step 7: Input the operating characteristics of the infusion pump motor to be tested into the motor fault detection SVM model, and output the fault detection results of the infusion pump motor to be tested according to the classification hyperplane parameters.

[0012] Support Vector Machine (SVM) is a data mining method based on statistical learning theory. Its mechanism involves finding a hyperplane that satisfies classification requirements, theoretically enabling optimal classification of linearly separable data. For infusion pump motors, various types of faults exist, such as rotational faults, noise faults, and temperature faults, each with varying frequencies and severity. To enable the model to learn the boundaries of critical faults, this invention sets different initial weights for different types of faults based on fault frequency and severity scores when defining the objective function of the motor fault detection SVM model. This results in different penalties for different fault types, forcing the model to effectively handle faults with low sample sizes. During subsequent training, the weights are adjusted inversely based on the misclassification loss, increasing the weights for persistently misclassified faults to optimize their classification boundaries. Finally, the objective function is optimized based on the updated weights, and a Lagrangian function is constructed and solved to obtain a decision boundary that maximizes the classification margin and minimizes the weighted error, improving the detection accuracy and adaptability of the motor fault detection SVM model.

[0013] Furthermore, in step one, the operating characteristics of the infusion pump motor include the motor output pressure, which is obtained through the following steps:

[0014] The measured pressure of the liquid in the pipeline is collected in real time using a pressure sensor;

[0015] Obtain the vertical height difference between the infusion pump and the patient's infusion site, and calculate the compensation pressure based on the vertical height difference and the drug density;

[0016] The motor output pressure is calculated based on the measured pressure and the compensation pressure.

[0017] In the training process of the motor fault detection SVM model, the model learns the differences in operating characteristics between normal and faulty motor conditions to distinguish different operating states. Motor output pressure, as a key operating characteristic, can reflect the motor's health status along with other features. However, during the use of infusion pumps, the measured pressure of the tubing fluid collected by pressure sensors often fails to reflect the true motor output pressure. This is because the height difference between the infusion pump and the patient's infusion site causes a certain static pressure in the infusion tubing, which is actually included in the measured pressure. This invention obtains the vertical height difference between the infusion pump and the patient's infusion site in real time, and then uses this vertical height difference and the drug density to calculate the compensation pressure, ultimately obtaining an accurate motor output pressure. This provides more reliable data input for training the motor fault detection SVM model.

[0018] Furthermore, in step one, the operating characteristics of the infusion pump motor also include: effective ripple characteristics of the motor phase current, which are obtained through the following steps:

[0019] The dynamic changes in the motor output pressure are monitored. When the motor output pressure changes abruptly, the rate of change of the motor output pressure over time during the abrupt change period is calculated, and a pressure change rate sequence is generated.

[0020] Synchronously collect the current waveform data of the infusion pump motor winding during the abrupt change period, calculate the current fluctuation rate in each driving cycle based on the current waveform data, and generate a current fluctuation rate sequence.

[0021] Calculate the time-domain correlation coefficient between the pressure change rate sequence and the current fluctuation rate sequence. If the time-domain correlation coefficient exceeds a preset coefficient threshold, then mark the current fluctuation rate sequence as the effective ripple feature of the motor phase current.

[0022] The infusion pump uses a stepper motor, which may experience micro-stepping faults under low speed and sudden load changes. Micro-stepping can lead to inaccurate drug delivery, affecting infusion accuracy and causing adverse consequences. This type of fault is typically overlooked by traditional motor detection models. This invention focuses on the coupling characteristics between current fluctuation rate and motor output pressure. By analyzing the time-domain correlation between the pressure change rate sequence and the current fluctuation rate sequence, it identifies the effective ripple characteristics of the motor phase current caused by pressure disturbances in the liquid pipeline and uses this as a feature dimension for micro-stepping faults, thus improving the sensitivity of the motor fault detection SVM model to early-stage stepping faults.

[0023] Furthermore, in step one, the operating characteristics of the infusion pump motor also include: dynamic characteristics of the patient's infusion site, which are obtained through the following steps:

[0024] Real-time acquisition of the duration of displacement of the patient's infusion site and the three-dimensional spatial displacement acceleration of the patient's infusion site;

[0025] The instantaneous displacement of the patient's infusion site is obtained by integrating the three-dimensional spatial displacement acceleration.

[0026] If the displacement duration is lower than a preset time threshold and the instantaneous displacement exceeds a preset displacement threshold, then the displacement duration and the instantaneous displacement are marked as dynamic features of the patient's infusion site.

[0027] Sudden changes in patient position during intravenous infusion, such as abruptly raising or lowering an arm, or turning over, can cause significant traction on the infusion tubing, leading to slight vibrations in the infusion pump. These vibrations, detected by a vibration sensor, may be misinterpreted as motor malfunctions. Therefore, this invention identifies such interference by detecting the dynamic characteristics of the patient's infusion site. Specifically, when the displacement duration is below a preset time threshold and the instantaneous displacement exceeds a preset displacement threshold, it indicates that the patient's infusion site has undergone a positional change that could cause pump vibration within a short period. In this case, the displacement duration and instantaneous displacement are used as dynamic feature inputs of the patient's infusion site, enabling the motor fault detection SVM model to identify such vibrations and improve detection accuracy.

[0028] Furthermore, in step two, the calculation of initial weights for different types of faults based on fault frequency and fault severity scores includes:

[0029] Failure frequency is calculated based on historical records;

[0030] The severity of the fault is assessed based on the level of threat to patient safety, the level of maintenance cost of the infusion pump motor, and the level of downtime, resulting in the fault severity score.

[0031] The initial weight of the fault is calculated using the following formula:

[0032]

[0033] in, Let F be the initial weight for the i-th type of fault. i S is the fault frequency of type i fault. i The severity score for the i-th type of fault.

[0034] Among these, the three dimensions—the threat level to patient safety, the maintenance cost level of the infusion pump motor, and the downtime level—can comprehensively and accurately reflect the severity of a certain type of fault. Calculating initial weights for different types of faults based on fault frequency and severity scores can balance the impact of uneven dataset distribution. For example, faults with low frequency but high severity can also receive initial weights, prompting the classifier to pay more attention to these faults during training, thereby improving recognition accuracy.

[0035] Furthermore, in step three, the objective function is:

[0036]

[0037] Where w is the weighting coefficient, b is the bias, and ξ is the weighting coefficient. i Let x be a slack variable, C be the regularization parameter, n be the number of samples, and x be the regularization parameter. i Let y be the feature vector of the i-th type of fault. i For the output value of the i-th type of fault, w·x i +b is the equation of the classification hyperplane.

[0038] The objective function comprises two parts: maximizing the margin term (the larger this term, the stronger the model's generalization ability); and minimizing the loss term (this part aims to allow some data points to violate the margin constraint to a certain extent, thus handling noise and outliers). These two parts are balanced by a regularization parameter to ensure a balance between accuracy and generalization ability. Combining these two parts, the SVM objective function can be expressed as a constrained optimization problem. By solving this problem, the optimal classification hyperplane parameters w and b can be found.

[0039] Furthermore, in step four, the formula for calculating the misclassification loss of fault detection is:

[0040] L i =max(0,1-y) i (w·x i +b));

[0041] Among them, L i Let be the misclassification loss for the i-th type of fault.

[0042] Furthermore, in step five, the formula for updating the weights of the corresponding fault types based on the misclassification loss is:

[0043]

[0044] Where η is the learning rate, t is the number of iterations, and α iLet be the weight of the i-th type of fault. Misclassification loss refers to the loss value assigned to samples that are misclassified during the classification process by the SVM model. By minimizing the misclassification loss, the model can continuously adjust its parameters to better adapt to the training data, thereby improving classification accuracy.

[0045] Furthermore, in step six, the constructed Lagrange function is:

[0046]

[0047] Where, λ i These are Lagrange multipliers. By constructing a Lagrange function, the objective function and constraints can be integrated into a single function, thus simplifying the problem-solving process.

[0048] Furthermore, in step seven, the fault decision classification expression for outputting the fault detection result of the infusion pump motor to be tested based on the classification hyperplane parameters is as follows:

[0049]

[0050] Where, k(x) i ,x) is the kernel function of the first motor fault detection SVM model. γ is the kernel function parameter, and x is the feature vector of the infusion pump motor to be tested.

[0051] One or more technical solutions provided by this invention have at least the following technical effects or advantages:

[0052] 1. This invention assigns different initial weights to different types of faults based on fault frequency and fault severity scores, enabling the model to effectively handle imbalanced fault samples in infusion pump motors and improve the classification accuracy of critical faults. This invention employs an adaptive weighting mechanism, adjusting weights based on misclassification losses during model training to optimize the model and improve its generalization ability.

[0053] 2. This invention addresses the micro-stepping problem caused by sudden changes in pipeline pressure when the infusion pump motor is running at low speed. It innovatively adopts the effective ripple characteristics of the motor phase current as its exclusive detection dimension, thereby improving the sensitivity of the motor detection SVM model to the micro-stepping fault of the motor.

[0054] 3. This invention reflects changes in the patient's position during infusion by detecting the dynamic characteristics of the infusion site, preventing infusion pump vibration caused by the patient's unintentional pulling on the infusion tubing from being diagnosed as a motor malfunction, thus improving the accuracy of model detection. Attached Figure Description

[0055] The accompanying drawings, which are provided to further illustrate embodiments of the invention and constitute a part of this invention, are not intended to limit the scope of the invention.

[0056] Figure 1 This is a flowchart illustrating a fault detection method for an infusion pump motor based on an adaptive weighted SVM model, as described in this invention. Detailed Implementation

[0057] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, where there is no conflict, the embodiments of the present invention and the features thereof can be combined with each other.

[0058] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0059] Example 1

[0060] Please refer to Figure 1 Embodiment 1 of the present invention provides a method for fault detection of infusion pump motor based on an adaptive weight SVM model, the method comprising the following steps:

[0061] Step 1: Obtain historical data on the operating characteristics of the infusion pump motor and the corresponding operating status of the infusion pump motor to obtain a sample set;

[0062] Step 2: Calculate the initial weights for different types of faults based on fault frequency and fault severity scores;

[0063] Step 3: Construct a motor fault detection SVM model and define an objective function based on the initial weights;

[0064] Step 4: Train the motor fault detection SVM model using the sample set and calculate the misclassification loss of fault detection;

[0065] Step 5: Update the weights of the corresponding fault types based on the misclassification loss;

[0066] Step 6: Optimize the objective function based on the updated weights, construct the Lagrangian function and solve for it to obtain the classification hyperplane parameters;

[0067] Step 7: Input the operating characteristics of the infusion pump motor to be tested into the motor fault detection SVM model, and output the fault detection results of the infusion pump motor to be tested according to the classification hyperplane parameters.

[0068] In step one, the operating characteristics of the infusion pump motor can include conventional operating characteristics, such as pipeline flow rate, motor voltage, current, and other parameters, as well as the statistical characteristics of these parameters, such as mean, variance, skewness, and kurtosis, and frequency domain characteristics, such as power spectral density. Those skilled in the art can select these characteristics according to the actual situation. In this embodiment, data preprocessing is performed before feature extraction, including filtering high-frequency noise and normalization. The operating status of the infusion pump motor includes fault status and healthy status.

[0069] In this embodiment, in step one, the operating characteristics of the infusion pump motor include the motor output pressure, which is obtained through the following steps:

[0070] The measured pressure of the liquid in the pipeline is collected in real time using a pressure sensor;

[0071] Obtain the vertical height difference between the infusion pump and the patient's infusion site, and calculate the compensation pressure based on the vertical height difference and the drug density;

[0072] The motor output pressure is calculated based on the measured pressure and the compensation pressure.

[0073] The vertical height difference between the infusion pump and the patient's infusion site can be measured using existing computer vision technology. This involves first capturing images of the infusion pump and the patient's infusion site using a camera, identifying the images and obtaining their three-dimensional coordinates, and then calculating the vertical height difference based on these coordinates. The formula for calculating the compensation pressure is as follows:

[0074] P 补偿 =ρgh;

[0075] Among them, P 补偿 To compensate for the pressure, ρ is the density of the infusion fluid, g is the acceleration due to gravity, and h is the vertical height difference between the infusion pump and the patient's infusion site. The motor output pressure is obtained by subtracting the compensation pressure from the measured pressure.

[0076] In this embodiment, step one, the operating characteristics of the infusion pump motor further includes: effective ripple characteristics of the motor phase current, which are obtained through the following steps:

[0077] The dynamic changes in the motor output pressure are monitored. When the motor output pressure changes abruptly, the rate of change of the motor output pressure over time during the abrupt change period is calculated, and a pressure change rate sequence is generated.

[0078] Synchronously collect the current waveform data of the infusion pump motor winding during the abrupt change period, calculate the current fluctuation rate in each driving cycle based on the current waveform data, and generate a current fluctuation rate sequence.

[0079] Calculate the time-domain correlation coefficient between the pressure change rate sequence and the current fluctuation rate sequence. If the time-domain correlation coefficient exceeds a preset coefficient threshold, then mark the current fluctuation rate sequence as the effective ripple feature of the motor phase current.

[0080] Specifically, when the change in motor output pressure exceeds a preset pressure change threshold within a preset abrupt change time threshold, a sudden change in motor output pressure is determined to have occurred. At this time, the rate of change of motor output pressure over time is recorded simultaneously. For example, if a sudden change in motor output pressure is detected at a certain moment, the rate of change of motor output pressure relative to a fixed time interval Δt during the abrupt change period is recorded, generating a pressure change rate series. Wherein, ΔP is the pressure change value within a fixed time interval Δt. N is an integer.

[0081] Current fluctuation rate is calculated as follows: Where δ is the current fluctuation rate, and ΔI is the sum of the absolute values ​​of the differences between adjacent current peaks and troughs in the infusion pump motor windings within each motor drive cycle. For example, if three peak-trough pairs are measured in a certain drive cycle during a sudden change period: 0.8A→0.2A, 0.7A→0.15A, 0.75A→0.18A, then ΔI==|0.8-0.2|+|0.7-0.15|+|0.75-0.18|=1.67A, I A This represents the average current magnitude for this period. The current volatility sequence during the abrupt change period is as follows: Where n is the number of motor drive cycles during the abrupt change period.

[0082] The time-domain correlation coefficient can be selected from Pearson correlation coefficient or Spearman rank correlation coefficient, etc., and the preset coefficient threshold can be set by those skilled in the art according to actual conditions.

[0083] In this embodiment, step one, the operating characteristics of the infusion pump motor further includes: dynamic characteristics of the patient's infusion site, which are obtained through the following steps:

[0084] Real-time acquisition of the duration of displacement of the patient's infusion site and the three-dimensional spatial displacement acceleration of the patient's infusion site;

[0085] The instantaneous displacement of the patient's infusion site is obtained by integrating the three-dimensional spatial displacement acceleration.

[0086] If the displacement duration is lower than a preset time threshold and the instantaneous displacement exceeds a preset displacement threshold, then the displacement duration and the instantaneous displacement are marked as dynamic features of the patient's infusion site.

[0087] The three-dimensional spatial displacement acceleration of the patient's infusion site can be measured using an inertial measurement unit (IMU) fixed next to the infusion site. The displacement duration can be calculated using a timestamp attached to the IMU. When calculating the instantaneous displacement of the patient's infusion site based on the integral of the three-dimensional spatial displacement acceleration, the acceleration data is first integrated once to obtain velocity data. Since acceleration data is usually discrete, this embodiment uses numerical integration methods (such as the trapezoidal rule, Simpson's rule, etc.) for calculation. Then, the velocity data is integrated a second time to obtain displacement data, finally yielding the instantaneous displacement. Those skilled in the art can set preset time and displacement thresholds according to actual conditions; this invention does not limit this.

[0088] In step two, the calculation of initial weights for different types of faults based on fault frequency and fault severity scores includes:

[0089] Failure frequency is calculated based on historical records;

[0090] The severity of the fault is assessed based on the level of threat to patient safety, the level of maintenance cost of the infusion pump motor, and the level of downtime, resulting in the fault severity score.

[0091] The initial weight of the fault is calculated using the following formula:

[0092]

[0093] in, Let F be the initial weight for the i-th type of fault. i S is the fault frequency of type i fault. i The severity score for the i-th type of fault.

[0094] The failure frequency can be calculated by dividing the number of failures by the total running time.

[0095] The threat level of a malfunction to patient safety can be graded from Level 1 (no threat) to Level 5 (extreme threat) based on factors such as the degree of patient harm that the malfunction may cause and the severity of treatment interruption. Risk assessment tools and methods, such as FMEA and HAZOP, can be used to assist in the assessment.

[0096] The maintenance cost of an infusion pump motor can be calculated by considering factors such as the cost of spare parts and labor required for maintenance, as well as the model and manufacturer of the infusion pump.

[0097] Finally, the severity score of the fault is calculated using a weighted average method.

[0098] For example, suppose there are two types of faults: the first type of fault has a failure frequency of 4 times per 1000 hours, and the second type of fault has a failure frequency of 10 times per 1000 hours.

[0099] The assessment determined that the first type of fault posed a threat level of 5 to patient safety, a maintenance cost level of the infusion pump motor, and a downtime level of 4, respectively. The second type of fault was assessed as having a level of 3, a level of 1, and a level of 2, respectively.

[0100] First, calculate the severity score of the fault. Assuming that the threat level of the fault to patient safety is assigned a weight of 0.5, and the maintenance cost level and downtime level of the infusion pump motor are assigned weights of 0.25 respectively, then the weighted severity score of the first type of fault is 4, and the severity score of the second type of fault is 2.25.

[0101] Then, by combining the fault frequency and fault severity scores, the initial weights of the first type of fault and the second type of fault were calculated to be 0.16 and 0.225, respectively, and after normalization, they were 0.42 and 0.58, respectively.

[0102] In step three, the objective function is:

[0103]

[0104] Where w is the weighting coefficient, b is the bias, and ξ is the weighting coefficient. i Let x be a slack variable, C be the regularization parameter, n be the number of samples, and x be the regularization parameter. i Let y be the feature vector of the i-th type of fault. i For the output value of the i-th type of fault, w·x i +b is the equation of the classification hyperplane.

[0105] In step four, the formula for calculating the misclassification loss of fault detection is as follows:

[0106] L i =max(0,1-y) i (w·x i +b));

[0107] Among them, L i Let be the misclassification loss for the i-th type of fault.

[0108] In step five, the formula for updating the weights of the corresponding fault types based on the misclassification loss is as follows:

[0109]

[0110] Where η is the learning rate, t is the number of iterations, and α i Let be the weight of the i-th type of fault.

[0111] In step six, the constructed Lagrange function is:

[0112]

[0113] Where, λ i It is a Lagrange multiplier.

[0114] In step seven, the fault decision classification expression for outputting the fault detection result of the infusion pump motor to be tested based on the classification hyperplane parameters is as follows:

[0115]

[0116] Where, k(x) i ,x) is the kernel function of the first motor fault detection SVM model. γ is the kernel function parameter, which is obtained through cross-validation in this embodiment. x is the feature vector of the infusion pump motor to be tested. When f(x) = 0, it indicates that the infusion pump motor is currently in a healthy state; when f(x) = 1, it indicates that the infusion pump motor is currently in a fault state.

[0117] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0118] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for fault detection of an infusion pump motor based on an adaptive weighted SVM model, characterized in that, The method includes the following steps: Step 1: Obtain historical data on the operating characteristics of the infusion pump motor and the corresponding operating status of the infusion pump motor to obtain a sample set; Step 2: Calculate the initial weights for different types of faults based on fault frequency and fault severity scores; Step 3: Construct a motor fault detection SVM model and define an objective function based on the initial weights; Step 4: Train the motor fault detection SVM model using the sample set and calculate the misclassification loss of fault detection; Step 5: Update the weights of the corresponding fault types based on the misclassification loss; Step 6: Optimize the objective function based on the updated weights, construct the Lagrangian function and solve for it to obtain the classification hyperplane parameters; Step 7: Input the operating characteristics of the infusion pump motor to be tested into the motor fault detection SVM model, and output the fault detection results of the infusion pump motor to be tested according to the classification hyperplane parameters; In step one, the operating characteristics of the infusion pump motor include the motor output pressure and the effective ripple characteristics of the motor phase current. The motor output pressure is obtained through the following steps: The measured pressure of the liquid in the pipeline is collected in real time using a pressure sensor; Obtain the vertical height difference between the infusion pump and the patient's infusion site, and calculate the compensation pressure based on the vertical height difference and the drug density; The motor output pressure is calculated based on the measured pressure and the compensation pressure. The effective ripple characteristics of the motor phase current are obtained through the following steps: The dynamic changes in the motor output pressure are monitored. When the motor output pressure changes abruptly, the rate of change of the motor output pressure over time during the abrupt change period is calculated, and a pressure change rate sequence is generated. Synchronously collect the current waveform data of the infusion pump motor winding during the abrupt change period, calculate the current fluctuation rate in each driving cycle based on the current waveform data, and generate a current fluctuation rate sequence. Calculate the time-domain correlation coefficient between the pressure change rate sequence and the current fluctuation rate sequence. If the time-domain correlation coefficient exceeds a preset coefficient threshold, then mark the current fluctuation rate sequence as the effective ripple feature of the motor phase current.

2. The method for fault detection of an infusion pump motor based on an adaptive weighted SVM model according to claim 1, characterized in that, In step one, the operating characteristics of the infusion pump motor also include: dynamic characteristics of the patient's infusion site, which are obtained through the following steps: Real-time acquisition of the duration of displacement of the patient's infusion site and the three-dimensional spatial displacement acceleration of the patient's infusion site; The instantaneous displacement of the patient's infusion site is calculated based on the three-dimensional spatial displacement acceleration. If the displacement duration is lower than a preset time threshold and the instantaneous displacement exceeds a preset displacement threshold, then the displacement duration and the instantaneous displacement are marked as dynamic features of the patient's infusion site.

3. The method for fault detection of an infusion pump motor based on an adaptive weighted SVM model as described in claim 1, characterized in that, In step two, the calculation of initial weights for different types of faults based on fault frequency and fault severity scores includes: Failure frequency is calculated based on historical records; The severity of the fault is assessed based on the level of threat to patient safety, the level of maintenance cost of the infusion pump motor, and the level of downtime, resulting in the fault severity score. The initial weight of the fault is calculated using the following formula: ； in, Let be the initial weights for the i-th type of fault. Let be the fault frequency of type i. The severity score for the i-th type of fault.

4. The method for fault detection of an infusion pump motor based on an adaptive weighted SVM model according to claim 3, characterized in that, In step three, the objective function is: ； in, These are the weighting coefficients. For bias, As slack variables, For regularization parameters, For the sample size, Let i be the feature vector of the i-th type of fault. For the i-th type of fault, the output value is... For classifying hyperplane equations.

5. The method for fault detection of an infusion pump motor based on an adaptive weighted SVM model according to claim 4, characterized in that, In step four, the formula for calculating the misclassification loss of fault detection is as follows: ； in, Let be the misclassification loss for the i-th type of fault.

6. The method for fault detection of an infusion pump motor based on an adaptive weighted SVM model according to claim 5, characterized in that, In step five, the formula for updating the weights of the corresponding fault types based on the misclassification loss is as follows: ； in, For learning rate, For the number of iterations, Let be the weight of the i-th type of fault.

7. The method for fault detection of an infusion pump motor based on an adaptive weighted SVM model according to claim 6, characterized in that, In step six, the constructed Lagrange function is: ； in, It is a Lagrange multiplier.

8. The method for fault detection of an infusion pump motor based on an adaptive weighted SVM model according to claim 7, characterized in that, In step seven, the fault decision classification expression for outputting the fault detection result of the infusion pump motor to be tested based on the classification hyperplane parameters is as follows: ； in, This is the kernel function of the motor fault detection SVM model. , For kernel function parameters, This is the feature vector of the infusion pump motor to be tested.

Images