A method, device, equipment and storage medium for diagnosing faults of satellite service computers

Abstract

The present invention discloses a method for diagnosing faults of a satellite service computer, comprising the following steps: obtaining characteristic parameters of a satellite service computer and performing normalization processing; inputting the normalized characteristic parameters into a fault diagnosis model for fault judgment; and implementing fault resolution measures according to a preset program based on the fault judgment result. A support vector machine algorithm in machine learning is used to establish and train characteristic parameters that can characterize the working state of the satellite service computer to obtain a fault diagnosis model, and the diagnosis model is pre-compiled into code and burned into the satellite service computer. After the satellite is launched into orbit, the satellite service computer runs software to perform real-time fault monitoring and implement fault resolution measures to prevent a single point fault from being processed and repaired in a timely manner and causing a chain reaction.

Claims

1. A method for diagnosing faults of satellite computer. It is characterized in that The following steps are involved: Based on the preset feature parameters, a fault diagnosis model is constructed through support vector machine; Using the characteristic data obtained in each state of normal operation of the satellite service computer as training data for the first label value of the fault diagnosis model; Using characteristic data of the satellite service computer in various possible abnormal working conditions as training data for the second label value of the fault diagnosis model; Burning and recording the trained fault diagnosis model into the satellite service computer; Obtain characteristic parameters of the satellite computer and perform normalization processing; Inputting the normalized characteristic parameters into a fault diagnosis model for fault diagnosis; Implement fault resolution measures according to preset procedures based on fault judgment results; The method for constructing a fault diagnosis model through a support vector machine based on preset characteristic parameters includes: using a Gaussian radial basis kernel function as a kernel function in a support vector machine algorithm, and the function formula is as follows: , where the parameter Value range 2 -10 to 2 5 ; Use the preset feature parameters as training samples, solve the convex optimization problem, obtain the classification model of the support vector machine, and store this model data in the Model structure; the structure contains the kernel function The six parts are as follows: value, total number of support vectors, dimension of characteristic parameters, constant term in decision function, support vector matrix and decision coefficient of support vector; variable names gamma, N, Z, b, X[i][j], a[i] are assigned to these six parts respectively; and the fault diagnosis model is obtained.

2. The method for diagnosing faults of a satellite service computer according to claim 1, It is characterized in that The method for obtaining characteristic parameters of a star service computer comprises: The hardware analog data of the satellite service computer is collected through an analog-to-digital conversion chip and a current collection chip; The software digital quantity data of the satellite computer is collected through the ECC management module.

3. The method for diagnosing faults of a satellite service computer according to claim 2, It is characterized in that The method of inputting the processed characteristic parameters into a fault diagnosis model for fault judgment comprises: Obtaining a sample array x_pre[Z] composed of the characteristic parameters; wherein the array contains Z-dimensional characteristic quantity sample data for fault diagnosis; Perform Z calculations in the inner loop to complete the kernel function calculation; The outer loop performs N calculations to complete the relationship between the input data and the support vectors; Output the result of the first label value or the second label value.

4. The method for diagnosing faults of a satellite service computer according to claim 1, It is characterized in that The method for performing normalization processing comprises: Obtaining the characteristic parameters; Using the linear normalization function: , where mean(x) is the average value of all values ​​in the dimension where x is located, max(x) is the maximum value of all values ​​in the dimension where x is located, and min(x) is the minimum value of all values. The feature parameters are normalized.

5. The method for diagnosing faults of a satellite service computer according to claim 1, It is characterized in that The implementation of the fault resolution measures according to the preset procedure based on the fault judgment result includes: Disconnect the unit module that is diagnosed as faulty; Restart the faulty unit module or start a unit module with the same function.

6. A fault diagnosis device for satellite computer, It is characterized in that The method for diagnosing a fault in a satellite service computer according to any one of claims 1 to 5 above comprises: The acquisition module is used to acquire characteristic parameters of the satellite service computer and perform normalization processing; before acquiring the characteristic parameters of the satellite service computer, a fault diagnosis model is constructed by a support vector machine based on preset characteristic parameters; characteristic data acquired in various states of normal operation of the satellite service computer is used as training data for the first label value of the fault diagnosis model; characteristic data of various possible abnormal working states of the satellite service computer is used as training data for the second label value of the fault diagnosis model; and the trained fault diagnosis model is burned and recorded into the satellite service computer; A judgment module, used for inputting the processed characteristic parameters into a fault diagnosis model for fault judgment; An execution module is used to implement fault resolution measures according to a preset procedure based on the fault judgment result; The method for constructing a fault diagnosis model through a support vector machine based on preset characteristic parameters includes: using a Gaussian radial basis kernel function as a kernel function in a support vector machine algorithm, and the function formula is as follows: , where the parameter Value range 2 -10 to 2 5 ; Use the preset feature parameters as training samples, solve the convex optimization problem, obtain the classification model of the support vector machine, and store this model data in the Model structure; the structure contains the kernel function The six parts are as follows: value, total number of support vectors, dimension of characteristic parameters, constant term in decision function, support vector matrix and decision coefficient of support vector; variable names gamma, N, Z, b, X[i][j], a[i] are assigned to these six parts respectively; and the fault diagnosis model is obtained.

7. An electronic device, It is characterized in that include: A memory, the memory being used to store a processing program; A processor, wherein when executing the processing program, the processor implements the method for diagnosing faults of a satellite service computer as described in any one of claims 1 to 5.

8. A readable storage medium, It is characterized in that The readable storage medium stores a processing program, and when the processing program is executed by a processor, the method for diagnosing faults of a satellite service computer as described in any one of claims 1 to 5 is implemented.

Images