A method for evaluating the health status of satellite control moment gyroscope on orbit

Abstract

The present invention discloses an on-orbit health status assessment method for a satellite control moment gyro, which selects on-orbit data of a single machine to be assessed as ideal data, extracts six elements of the control moment gyro, and calculates the expected value and standard deviation of each element, and uses the coefficient of variation method to obtain a normalized weight vector; a relative deviation matrix is ​​established based on the six elements and the ideal reference vector; and the weight vector and the relative deviation matrix are multiplied to obtain an assessment vector. The present invention uses known on-orbit data for analysis and statistics to find the inherent interdependence between multidimensional data in a stable and ideal operating state, thereby predicting abnormal trends, getting rid of the limitations of manual evaluation, discovering potential abnormalities and processing them in a timely manner, filling the current technical gap, and gaining time for product improvement and fault handling; using a data transformation calculation method, the six independent elements are organically unified, and a weighted quantitative comprehensive assessment is performed to achieve quantitative assessment, which is simple to calculate and easy to operate.

Description

Technical Field

[0001] The present invention relates to the technical field of on-orbit satellite health status assessment, and in particular to an on-orbit health status assessment method for a satellite control moment gyroscope. Background Art

[0002] CMG (Control Moment Gyro) is the key executive component for satellites to achieve agility. It consists of a high-speed rotor part and a low-speed frame part. The control system changes the angular momentum vector direction of each CMG to achieve angular momentum exchange between CMG and the spacecraft, thereby changing the attitude of the spacecraft. CMG can start and stop quickly and with high precision within a few seconds, allowing the entire satellite to maneuver in any direction, which is a qualitative leap compared to traditional momentum wheels. CMG adopts mechatronics design. The control circuit board, frame drive circuit board, power board and frame angle measurement board are all inside the product. The conductive ring transmits signals including high-speed wheel control, feedback and power supply, synchronizer rotor excitation, etc.

[0003] As a new product, CMG has been in orbit for a short time, but its number has increased rapidly. According to statistics, the performance of new CMG spectra has declined or abnormalities have occurred, or even become permanent faults, in an average of about two and a half years after they were put into use, which seriously affects their agility and life. At present, CMG on-orbit faults mainly include the following two categories: unstable contact inside the conductive ring, causing abnormal path impedance; differences in the wear resistance of the cage or accumulation of lubricating oil, resulting in increased friction torque of the bearing. For the first type of abnormality, the contact between the brush wire and the conductive ring is unstable, the impedance increases, and the instantaneous short circuit causes unstable signal transmission, no trend change abnormality, short-term sudden abnormality, and is eventually eliminated by the system. There is no trend change in the motor current, shaft temperature, and speed before and after the abnormality; for the second type of abnormality, the motor current and shaft temperature increase and the fluctuation gradually amplifies, with a trend change abnormality, but the time from the precursor to the final occurrence of the fault varies.

[0004] To maintain the healthy operation of a single machine, it is necessary to evaluate its performance, intervene in possible anomalies in advance, reduce the impact on the system, and provide a basis for system health management. The existing evaluation of CMG single-machine products mainly relies on human experience and knowledge, and manually interprets the trends of certain key parameters. With the accumulation of data and the sharp increase in the number of single machines on orbit, human limitations have become more and more prominent.

[0005] Therefore, how to find a comprehensive evaluation method to provide effective early warning for such sudden and slow-changing anomalies and get rid of the limitations of manual evaluation is an urgent need for the current CMG on-orbit performance evaluation. Summary of the invention

[0006] In view of this, the present invention provides an on-orbit health status assessment method for a satellite control moment gyro, which can get rid of the limitations of manual evaluation and effectively assess the health status of the control moment gyro. The assessment process is simple in calculation and easy to operate.

[0007] To achieve the above object, the technical solution of the present invention is:

[0008] A method for evaluating the on-orbit health status of a satellite control moment gyro is used to evaluate six elements of the control moment gyro. The specific steps include:

[0009] The on-orbit data of the single machine to be evaluated is selected as the ideal data, the six elements of the control moment gyro are extracted, and the expected value and standard deviation of each element are calculated; the expected value and standard deviation are substituted, and the normalized weight vector is obtained using the coefficient of variation method; based on the ideal data, an ideal reference vector is established; based on the six elements and the ideal reference vector, a relative deviation matrix is ​​established; the weight vector and the relative deviation matrix are multiplied to obtain the evaluation vector.

[0010] Furthermore, the weight vector The calculation formula is:

[0011]

[0012]

[0013] Among them, v j is the weight coefficient of the jth factor, which is the ratio of the standard deviation to the expected value; j is the number of factors, j=1,2,3,4,5,6.

[0014] Furthermore, an ideal reference vector is established based on the ideal data. The specific method is as follows:

[0015] Take the average value of the ideal data to form the ideal reference vector u:

[0016]

[0017] Among them, the ideal value of the i-th single machine is i is the number of single machines; x i is the six-element vector of the i-th machine.

[0018] Furthermore, a relative deviation matrix is ​​established based on the six element vectors and the ideal reference vector. The specific method is as follows:

[0019] Substituting the six-element vector and the ideal reference vector, we can get the relative deviation element r of the jth element of the i-th unit: ij :

[0020]

[0021] Among them, x ij is the jth element in the six-element vector of the i-th machine.

[0022] All relative deviation elements are combined into a relative deviation matrix R:

[0023]

[0024] Furthermore, the on-orbit data of the single machine to be evaluated is selected to be a period of 6-12 months.

[0025] Furthermore, the weight vector and the relative deviation matrix are multiplied to obtain the evaluation vector. The specific method is:

[0026] Evaluate the i-th element D in vector D i for:

[0027]

[0028] Furthermore, the six factors are high-speed motor current, high-speed rotor speed, high-speed rotor shaft temperature, low-speed motor current, low-speed rotor shaft temperature and frame angular velocity.

[0029] Beneficial effects:

[0030] 1. This method selects the on-orbit data of the single machine to be evaluated as the ideal data, extracts the six elements of the control moment gyro, and calculates the expected value and standard deviation of each element; substitutes the expected value and standard deviation, and uses the coefficient of variation method to obtain the normalized weight vector; establishes the ideal reference vector according to the ideal data; establishes the relative deviation matrix according to the six elements and the ideal reference vector; multiplies the weight vector and the relative deviation matrix to obtain the evaluation vector. The present invention uses known on-orbit data for analysis and statistics, finds the inherent interdependence between multidimensional data in the stable ideal operating state, thereby predicting abnormal trends, getting rid of the limitations of manual evaluation, discovering potential abnormalities and processing them in time, filling the gap that the performance degradation and abnormalities of the current on-orbit CMG products cannot be discovered early, and buying time for product improvement and fault handling; using the data transformation calculation method, the six independent elements are organically unified, and a weighted quantitative comprehensive evaluation is performed to achieve a quantitative evaluation of the on-orbit health status of CMG products, which is simple to calculate and easy to operate.

[0031] 2. The present invention comprehensively considers six factors that can reflect the on-orbit performance of CMG, and normalizes the weight coefficients (weight vectors) of the six factors and the deviations from the reference values ​​(relative deviation matrices) based on the performance thresholds, so that the reconstructability comparison can be performed between physical quantities of different magnitudes and dimensions, which is more universal.

[0032] 3. In the present invention, if the ideal reference vector is taken from the average value of a set of data after stable on-orbit operation, all subsequent evaluation data will be used as the reference value. This method can obtain an absolute evaluation value, and can rank multiple devices configured on the same satellite by comprehensive scores to evaluate the relative quality of CMG products. If it is taken from the average value of each evaluation result itself, this method will obtain a relative evaluation value. By comparing the differences, the equipment that needs to be focused on can be determined and applied to trend prediction.

[0033] 4. The evaluation method of the present invention has a certain degree of universality, and elements of other single-unit products of satellites (such as momentum wheels, gyroscopes, etc.) can be introduced to apply the method of the present invention to conduct on-orbit performance evaluation.

[0034] 5. The present invention can continuously improve the evaluation accuracy by continuously comparing the estimation results of the method with the actual situation and revising the evaluation model. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 The present invention is a flow chart of the method. DETAILED DESCRIPTION

[0036] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0037] The parameters of the high-speed performance of CMG (Control Moment Gyro) mainly include motor current, high-speed rotor speed, and high-speed rotor shaft temperature. The parameters that characterize the low-speed performance of CMG mainly include: motor current, low-speed rotor shaft temperature, frame angular velocity, involving different physical quantities such as electricity, heat, and angular velocity.

[0038] The solution of the present invention is: based on a large amount of on-orbit telemetry data, the present invention proposes a comprehensive evaluation method based on data transformation, comprehensively considers the on-orbit characteristics of the product, selects evaluation factors, and uses the coefficient of variation method to establish a normalized weight vector of each factor, compares the actual on-orbit data with the ideal value to obtain a relative deviation matrix, and finally establishes an on-orbit comprehensive scoring model for CMG products to achieve a quantitative evaluation of the on-orbit performance of CMG products. This method is used to conduct on-orbit health evaluation of CMG products of multiple satellites. The results show that the recognition rate of abnormal trends in CMG on-orbit performance is high, and it can be applied to on-orbit health evaluation and trend warning of similar products.

[0039] Based on the data transformation comprehensive evaluation model, the above parameters of CMG are analyzed. The parameters of the high-speed performance of CMG are mainly: motor current, high-speed rotor speed, and high-speed rotor shaft temperature. The parameters that characterize the low-speed performance of CMG are mainly: motor current, low-speed rotor shaft temperature, and frame angular velocity. The index parameters include amplitude and fluctuation. In the present invention, the six-element matrix is:

[0040]

[0041] Where: x i ={x i1 ,x i2 ,x i3 ,x i4 ,x i5 ,x i6} respectively represent the high-speed motor current, high-speed rotor speed, high-speed rotor shaft temperature, low-speed motor current, low-speed rotor shaft temperature, and frame angular velocity of the ith unit (the six-element vector of the ith unit). The six-element vectors of all units constitute a six-element matrix.

[0042] In fact, the performance degradation or failure of a single machine is often not caused by a serious over-limit of a single indicator, but by the degradation of multiple indicators, and the coupling effect causes system failure. Therefore, in order to identify and distinguish the impact of multiple indicators, it is necessary to design a method to quantify the weight of each indicator.

[0043] like Figure 1 As shown, the present invention provides an on-orbit health status assessment method for a satellite control moment gyro, which is used to assess six elements of the control moment gyro (high-speed motor current, high-speed rotor speed, high-speed rotor shaft temperature, low-speed motor current, low-speed rotor shaft temperature and frame angular velocity), and the specific steps include:

[0044] The on-orbit data of the single aircraft to be evaluated is selected as the ideal data, the six elements of the control moment gyro are extracted, and the expected value and standard deviation of each element are calculated. The on-orbit data of the single aircraft to be evaluated is selected from a period of 6-12 months.

[0045] Substitute the expected value and standard deviation and use the coefficient of variation method to get the normalized weight vector. The calculation formula is:

[0046]

[0047]

[0048] Among them, v j is the weight coefficient of the jth factor, which is the ratio of the standard deviation to the expected value; j is the number of factors, j = 1, 2, 3, 4, 5, 6. The advantage of this step is that these indicators with different physical meanings are unified into one evaluation system for normalized calculation.

[0049] During the system evaluation process, it is necessary to evaluate the degree of deviation of each indicator. Therefore, this method needs to give an ideal reference vector, and evaluate the degree of deviation of each indicator from the ideal value through a certain comparison method. The ideal reference vector can be given based on theoretical values, engineering experience, or statistical methods based on actual on-orbit data. Considering the different orbits, application scenarios of satellites, and the technical iteration of CMG itself, theoretical values ​​and empirical values ​​are difficult to obtain. In this method, the ideal reference vector of CMG gives the telemetry data baseline value using the on-orbit data statistics method. The advantage of this method is that the ideal value can be dynamically iterated and corrected according to the actual on-orbit operation conditions. Based on the ideal data, an ideal reference vector is established. The average value of the ideal data is taken to form the ideal reference vector u:

[0050]

[0051] Among them, the ideal value of the i-th single machine is i is the number of single machines; x i is the six-element vector of the ith single machine. The advantage of this step is that when the sample data is not abundant, it does not rely on theoretical values, and only selects a section of data with good operating status as the ideal reference vector by a priori on-orbit.

[0052] The six-dimensional evaluation indicators selected in this method involve different physical quantities such as electricity, heat, and speed, and the goal of product performance evaluation is to unify these indicators with different physical meanings into an evaluation system. Therefore, in this step, in order to integrate all indicators, this method establishes a relative deviation matrix based on the six elements and the ideal reference vector. First, substitute the six element vectors and the ideal reference vector to obtain the relative deviation element r of the jth element of the i-th unit. ij .

[0053]

[0054] Among them, x ij is the jth element in the six-element vector of the i-th machine. It represents the average value of the jth evaluation factor of the ith single machine after removing outliers during the target evaluation period; max{x ij} represents the maximum value of the jth evaluation factor of the ith single machine after eliminating outliers; min{x ij The minimum value of the jth evaluation factor of the ith single machine after eliminating outliers. The key point of this step is to change the absolute value into a relative value through the difference between the target data and the ideal data, and then normalize the relative values ​​of the six evaluation target parameters by dividing with the extreme value difference.

[0055] All relative deviation elements are combined into a relative deviation matrix R:

[0056]

[0057] This method first determines the weight coefficient of each indicator, and then establishes a normalized relative deviation matrix based on the ideal reference vector. The weight vector and the relative deviation matrix are multiplied to obtain the evaluation vector. The specific method is:

[0058] Evaluate the i-th element D in vector D i for:

[0059]

[0060] Furthermore, if the ideal reference vector is taken from the average value of a set of data after stable on-orbit operation, all subsequent evaluation data will be used as the reference value. This method will obtain an absolute evaluation value, and multiple devices configured with the same satellite will be ranked by comprehensive scores to evaluate the relative quality of CMG products. If it is taken from the average value of each evaluation result itself, this method will obtain a relative evaluation value, and the equipment that needs to be focused on will be determined through difference comparison. Therefore, the present invention can also be applied to trend prediction.

[0061] An embodiment is given below to further illustrate the present invention.

[0062] Step 1: Select a piece of ideal data and calculate the expected value and standard deviation of each factor.

[0063] A period of data on the stable operation of the six elements TMK01 (high-speed motor current), TMK02 (high-speed speed), TMK03 (high-speed shaft temperature), TMK04 (low-speed motor current), TMK05 (low-speed bearing temperature), and TMK06 (low-speed frame angular velocity) is extracted from the historical telemetry database of the in-orbit satellite. The data should not be too short, and 6-12 months is preferred.

[0064] After eliminating obviously unreasonable outlier data, the expected value and standard deviation of the extracted six factors were calculated.

[0065] For example, there are four main types of CMGs currently on orbit, namely 5Nms, 15Nms, 25Nms and 200Nms. The telemetry data of stable operation from June 1, 2016 to June 1, 2017 was selected to obtain the expected value and standard value of each indicator, as shown in Table 1. The detailed implementation steps are as follows.

[0066] Table 1 Statistics of standard values ​​and expected values ​​of parameters of four control moment gyros

[0067]

[0068]

[0069] Step 2: Calculate the weight coefficient and ideal benchmark value.

[0070] Using the formula The weight coefficients of each indicator of each type of CMG will be calculated to obtain the weight coefficient matrix. For example, the weight coefficients obtained from the data from June 1, 2016 to June 1, 2017 are:

[0071]

[0072] Create an ideal benchmark value matrix:

[0073] in:

[0074] The ideal values ​​of each indicator are averaged to obtain the following ideal benchmark value matrix:

[0075]

[0076] So far, this embodiment uses the ideal data from June 1, 2016 to June 1, 2017 as a sample to obtain the weight coefficients and ideal benchmark values ​​of the six indicators of various types of CMG.

[0077] Step 3: Select telemetry data for the target evaluation period.

[0078] The following is an example of a satellite with 5 15Nms control torque gyros. The telemetry data from April 1, 2017 to April 1, 2018 are taken to calculate the average value of each item:

[0079]

[0080] Substitute into the formula

[0081] where u 15N =(2.1665 6000.453 2.784 3.1375 2.412 0.06745), and we get the offset matrix:

[0082]

[0083] The weight coefficient vector is:

[0084] w 15N =[0.00075326 0.00000022 0.00016904 0.00057131 0.000216400.99828977]

[0085] Using the formula: Thus, the final evaluation scores of each CMG are obtained:

[0086] D1=0.003038694

[0087] D2=0.004784951

[0088] D3=0.003159483

[0089] D4=0.002595248

[0090] D5=0.001321537

[0091] If D i <D j , then CMG i Rating is better than CMG j Therefore, the evaluation ranking of the satellite's five control moment gyros is:

[0092] CMG5>CMG4>CMG1>CMG3>CMG2.

[0093] In summary, the above are only preferred embodiments of the present invention and are not intended to limit the protection scope of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for evaluating the on-orbit health status of a satellite control moment gyro, characterized in that: The six elements used to evaluate the control moment gyro include: Select the on-orbit data of the single machine to be evaluated as the ideal data, extract the six elements of the control moment gyro, and calculate the expected value and standard deviation of each element; substitute the expected value and standard deviation, and use the coefficient of variation method to obtain the normalized weight vector; establish the ideal reference vector based on the ideal data; establish the relative deviation matrix based on the six elements and the ideal reference vector; multiply the weight vector and the relative deviation matrix to obtain the evaluation vector; The specific method is: Take the average value of the ideal data to form the ideal reference vector u: Among them, the ideal value of the i-th single machine is i is the number of single machines; x i is the six-element vector of the i-th single machine; The relative deviation matrix is ​​established according to the six elements and the ideal reference vector. The specific method is as follows: Substituting the six-element vector and the ideal reference vector, we can get the relative deviation element r of the jth element of the i-th unit: ij : Among them, x ij is the jth element in the six-element vector of the i-th single machine; All relative deviation elements are combined into a relative deviation matrix R: The six elements are high-speed motor current, high-speed rotor speed, high-speed rotor shaft temperature, low-speed motor current, low-speed rotor shaft temperature and frame angular velocity.

2. The method according to claim 1, characterized in that The weight vector The calculation formula is: Among them, the v j is the weight coefficient of the jth factor, which is the ratio of the standard deviation to the expected value; j is the number of factors, j = 1, 2, 3, 4, 5, 6.

3. The method according to claim 1, characterized in that The on-orbit data of the single machine to be evaluated is selected from a period of 6-12 months.

4. The method according to claim 1, characterized in that The weight vector and the relative deviation matrix are multiplied to obtain the evaluation vector. The specific method is as follows: Evaluate the i-th element D in vector D i for:

Images