A performance status evaluation method for inertial navigation system considering non-periodic data

Abstract

The present application discloses a method for evaluating the performance status of an inertial navigation system taking into account non-periodic data, and relates to the field of health management of inertial navigation systems. The method includes establishing a performance evaluation model for an inertial navigation system; combining the lateral weights and longitudinal weights of each non-periodic test moment to obtain the comprehensive weights of each non-periodic test moment of the inertial navigation system; fusing the data of each historical non-periodic test moment with the data of the current test moment according to the comprehensive weight to generate test data; determining the weight of each indicator according to the comprehensive weight of each non-periodic test moment; determining the reliability of each indicator according to the test data; using the evidence reasoning algorithm to fuse each indicator in the test data according to the weight of each indicator and the reliability of each indicator, and using the performance evaluation model to obtain the performance status of the inertial navigation system at the current test moment. The present application can improve the performance evaluation accuracy of the inertial navigation system in an actual environment.

Description

Technical Field

[0001] The present application relates to the field of inertial navigation system health management, and in particular to a method for evaluating the performance status of an inertial navigation system taking into account non-periodic data. Background Art

[0002] The inertial navigation system provides continuous and accurate position and attitude information that does not rely on external signals by autonomously measuring motion parameters, which is crucial to ensuring the navigation safety and accuracy of its carrying equipment in various environments. Therefore, accurately grasping the performance status of the inertial navigation system is of great engineering significance.

[0003] The problems faced in the performance evaluation of inertial navigation systems can be summarized into three aspects: First, with the continuous improvement of the level of equipment manufacturing industry, the design reliability of inertial navigation systems has been continuously improved, and the probability of failure has been greatly reduced, which has led to a lack of abnormal information that can be obtained; Second, due to the high coupling between the various subsystems of the inertial navigation system and the use of integrated design, there are many factors that affect its working state. It is difficult to establish an accurate mathematical model based solely on expert knowledge. At the same time, the probability of abnormal state occurrence is greatly reduced, resulting in a lack of high-value abnormal information and an inability to provide an effective training tool for the model; Third, the test cycle of most inertial navigation systems is irregular. In addition to daily maintenance, performance testing is usually only performed before system maintenance or mission execution, which leads to the non-periodic characteristics of the historical data obtained and poses a challenge to its integration into the performance evaluation model at the current test moment. Therefore, failure to consider non-periodic historical testing will have an adverse effect on the performance evaluation of the inertial navigation system at the current test moment. Summary of the invention

[0004] The purpose of this application is to provide a method for evaluating the performance status of an inertial navigation system taking into account non-periodic data, which can improve the performance evaluation accuracy of the inertial navigation system in a practical environment.

[0005] To achieve the above objectives, this application provides the following solutions.

[0006] The present application provides a method for evaluating the performance status of an inertial navigation system considering non-periodic data, including: establishing a performance evaluation model of an inertial navigation system; determining the horizontal weight and the vertical weight of each non-periodic test moment of the inertial navigation system; all non-periodic test moments include multiple historical non-periodic test moments and a current test moment; combining the horizontal weight and the vertical weight of each non-periodic test moment to obtain the comprehensive weight of each non-periodic test moment of the inertial navigation system; according to the comprehensive weight of each non-periodic test moment, fusing the data of each historical non-periodic test moment with the data of the current test moment to generate test data; each moment in the test data corresponds to multiple indicators; according to the comprehensive weight of each non-periodic test moment, determining the weight of each indicator; according to the test data, determining the reliability of each indicator; according to the weight of each indicator and the reliability of each indicator, using an evidential reasoning algorithm to fuse each indicator in the test data, and using the performance evaluation model to obtain the performance status of the inertial navigation system at the current test moment.

[0007] According to the specific embodiments provided in this application, this application discloses the following technical effects.

[0008] The present application provides a method for evaluating the performance status of an inertial navigation system taking into account non-periodic data. The method combines the lateral weight and the longitudinal weight of each non-periodic test moment to obtain the comprehensive weight of each non-periodic test moment of the inertial navigation system. The data of each historical non-periodic test moment and the data of the current test moment are fused according to the comprehensive weight. Then, the performance status of the inertial navigation system at the current test moment is determined according to the fused non-periodic test data. This method effectively solves the problem that the performance evaluation accuracy of the inertial navigation system is affected by the non-periodic test data, and improves the performance evaluation accuracy of the inertial navigation system in an actual environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0010] Figure 1 A flowchart of a method for evaluating the performance status of an inertial navigation system taking into account non-periodic data is provided in accordance with an embodiment of the present application.

[0011] Figure 2 Schematic diagram of historical test data of the inertial navigation system used in the experiment.

[0012] Figure 3 Schematic diagram of the performance status of the inertial navigation system used in the experiment under historical non-periodic tests.

[0013] Figure 4 Schematic diagram of the current test data of the inertial navigation system used in the experiment.

[0014] Figure 5 This is a schematic diagram of the bidirectional weights during non-periodic testing.

[0015] Figure 6 Schematic diagram of comprehensive weights at non-periodic test moments.

[0016] Figure 7 Schematic diagram of 50 repeated experiments for the performance evaluation model of the inertial navigation system. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0018] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0019] In an exemplary embodiment, Figure 1 As shown, a method for evaluating the performance status of an inertial navigation system considering non-periodic data is provided, including the following steps 101 to 107.

[0020] Step 101: Establish a performance evaluation model of an inertial navigation system.

[0021] Step 102: Determine the lateral weight and longitudinal weight of each non-periodic test moment of the inertial navigation system; all non-periodic test moments include multiple historical non-periodic test moments and the current test moment.

[0022] Step 103: Combining the lateral weight and the longitudinal weight of each non-periodic test moment, obtain the comprehensive weight of each non-periodic test moment of the inertial navigation system.

[0023] Step 104: According to the comprehensive weight of each non-periodic test moment, the data of each historical non-periodic test moment and the data of the current test moment are integrated to generate test data; each moment in the test data corresponds to multiple indicators.

[0024] Step 105: Determine the weight of each indicator according to the comprehensive weight of each non-periodic test moment.

[0025] Step 106: Determine the reliability of each indicator based on the test data.

[0026] Step 107: According to the weight of each indicator and the reliability of each indicator, the evidential reasoning algorithm is used to fuse the indicators in the test data, and the performance evaluation model is used to obtain the performance status of the inertial navigation system at the current test moment.

[0027] By implementing the above steps 101 to 107, the comprehensive weight of each non-periodic test moment of the inertial navigation system is obtained by combining the lateral weight and the longitudinal weight of each non-periodic test moment. The data of each historical non-periodic test moment and the data of the current test moment are fused according to the comprehensive weight, and then the performance status of the inertial navigation system at the current test moment is determined according to the fused non-periodic test data, which effectively solves the problem that the performance evaluation accuracy of the inertial navigation system is affected by the non-periodic test data, and improves the performance evaluation accuracy of the inertial navigation system in the actual environment.

[0028] In another exemplary embodiment of the present application, according to the way of constructing evidence in the Evidential Reasoning (ER) algorithm, the performance evaluation model of the inertial navigation system in the above step 101 may be as follows.

[0029] .

[0030] In the formula, The performance evaluation model indicators; For the Considered in the index Data of non-periodic test moments, There are multiple performance states that may occur in the inertial navigation system. In order The corresponding possibility, For the The weight of the indicator, For the The reliability of the indicator, for A non-periodic test moment.

[0031] In another exemplary embodiment of the present application, the input information conversion of non-periodic historical data is first considered. Assume that The non-periodic data of the non-periodic test time are input at the following times: to , and presents a non-periodic form, the corresponding input is to ,in, to for A historical non-periodic test moment, is the current test time, The data at the current test time is used as the current input value. ,…, Represents all indicator input information, which is a vector. Different devices select different indicators. Taking the performance evaluation of the inertial navigation system as an example, the input information includes three accelerations and three gyroscopes' 0th-order drift coefficients, a total of six indicators. In order to make full use of these non-periodic data, this application assigns weights to the non-periodic historical data and combines it with the data at the current test moment into a new set of input values. The non-periodic data weight set is divided into horizontal weights and vertical weights: the horizontal weight is expressed in the time dimension, the closer to the current test moment, the larger the horizontal weight value; the vertical weight is in the dimension of device performance degradation, obtained according to the decline in performance utility in each time period, the more obvious the change, the larger the vertical weight. The calculation formula for the horizontal weight in the above step 102 is as follows.

[0032] .

[0033] In the formula, For the The horizontal weight of the non-periodic test moment, is the current test time, is the total number of non-periodic test moments, For the It can be seen that the calculation formula of the horizontal weight is a normal distribution.

[0034] The process of determining the longitudinal weight in the above step 102 can be replaced by the following steps 201 to 204.

[0035] Step 201: Establish a performance degradation degree expression: ; In the formula, For the The non-periodic test time and +1 The performance degradation during the time period between non-periodic test moments, For the The performance utility value of the inertial navigation system corresponding to the non-periodic test moment, For the +1 inertial navigation system performance utility value corresponding to the non-periodic test moment, For the The non-periodic test time and +1 time period between non-periodic test moments, The total number of non-periodic test moments.

[0036] The performance utility value at the historical non-periodic test moment is the performance evaluation result of the inertial navigation system tested previously, which is a known quantity. The performance utility value of the inertial navigation system evaluated at the current test moment is known at the next test moment.

[0037] Step 202: According to the performance degradation degree expression, the weight expression of the time period is determined as follows: ; In the formula, For the The non-periodic test time and +The weight of the time period between 1 non-periodic test moments, || is the absolute value symbol.

[0038] Considering the existence It is not easy to obtain the longitudinal weight when it is equal to 0. In this application, it is mapped into an exponential function, that is, the weight expression of the time period.

[0039] Step 203: According to the weight expression of the time period, the calculation formula for obtaining the longitudinal weight is: ; In the formula, For the The longitudinal weight of the non-periodic test moment, is the weight of the time period between the first non-periodic test moment and the second non-periodic test moment, For the +1 non-periodic test moment and +The weight of the time period between 2 non-periodic test moments.

[0040] Step 204: Obtain the longitudinal weight of the inertial navigation system at each non-periodic test moment according to the calculation formula of the longitudinal weight.

[0041] In another exemplary embodiment of the present application, the horizontal weight and vertical weight By considering them equally and performing normalization, bidirectional weight fusion can be achieved. Then the calculation formula of the comprehensive weight in the above step 103 can be the following formula.

[0042] .

[0043] In the formula, For the inertial navigation system The comprehensive weight of the non-periodic test moments, For the The horizontal weight of the non-periodic test moment, For the The longitudinal weight of the non-periodic test moment, The total number of non-periodic test moments.

[0044] In another exemplary embodiment of the present application, the generation of test data in the above step 104 can be replaced by the following steps 301 to 304.

[0045] Step 301: Based on the comprehensive weights of each non-periodic test moment, according to the formula , fuse the data of each historical non-periodic test time with the data of the current test time to obtain the fused non-periodic test data; where, is the fused non-periodic test data, is the data at the current test time, For the inertial navigation system The comprehensive weight of the non-periodic test moments, For the The data of a non-periodic test moment.

[0046] Step 302: When each non-periodic test moment corresponds to When there are 1 index, the fused non-periodic test data is expressed as: ;in, For input, is the input value of the first indicator in the fused non-periodic test data, is the first The input value of the indicator, is the first The input value of the indicator, is the confidence level of the first indicator, For the The confidence level of an indicator, For the The confidence level of an indicator.

[0047] Confidence represents the uncertainty of the input information. For example, (red, 0.9) means that the input has a 90% certainty of taking the value "red" of the attribute "color". In general, the value is 1.

[0048] Step 303: Convert each item in the fused non-periodic test data into a confidence distribution based on the index reference value: ; In the formula, For the The first indicator Results, for point to No. Reference values, for point to No. +1 reference value, is the number of results.

[0049] The main purpose of the form conversion in this step is to meet the input form requirements of the ER algorithm. The reference value refers to the numerical value of a certain indicator in various performance states, which is usually given by domain experts or equipment manuals.

[0050] Step 304: All confidence distribution forms based on the index reference values ​​constitute test data.

[0051] In another exemplary embodiment of the present application, the coefficient of variation is an important parameter for measuring the relative fluctuations between different data sets. When the coefficient of variation of a certain indicator is large, it indicates that its relative fluctuation is large, and a larger weight should be given to the indicator. Then the process of determining the weights of each indicator in the above step 105 can be replaced by the following steps 401 to 405.

[0052] Step 401: Establish a calculation formula for the first coefficient of variation: ; In the formula, For the The first coefficient of variation of an indicator at the current test moment; For the The average value of the indicators, For the The root mean square error of each indicator; The current test time.

[0053] and Both reflect the data discrete trend of the data set.

[0054] Step 402: Establish a calculation formula for the second coefficient of variation: ; In the formula, is the second coefficient of variation at the current test moment, For the inertial navigation system The comprehensive weight of the non-periodic test moments, is the comprehensive weighted average value of the inertial navigation system at all historical non-periodic test moments, The total number of non-periodic test moments.

[0055] The calculation of the second coefficient of variation also takes into account the impact of non-contemporary historical data on the indicator weight.

[0056] Step 403: According to the calculation formula of the first coefficient of variation and the calculation formula of the second coefficient of variation, the comprehensive coefficient of variation expression is determined as: ; In the formula, For the The comprehensive coefficient of variation of the indicators at the current test time, It is a dynamic adjustment coefficient.

[0057] Explain from the following three aspects .

[0058] 1) Although It can reflect the volatility of test data as a means of determining the relative importance of indicators, but it may be incomplete and inaccurate. , we can better integrate the volatility and time significance of the data, thus obtaining the coefficient of variation of the indicator.

[0059] 2) If If it is 1, then there is , which means that the weight is completely affected by the fluctuation of the test data. Therefore, the contribution of performance fluctuations in the time dimension to the indicator weight may be hindered.

[0060] 3) If If it is 0, then there is , which means that the weights are completely influenced by the time importance. Therefore, the contribution of volatility indicator weights at the data level may be hindered.

[0061] Step 404: Establish the calculation formula of the indicator weight: ; In the formula, For the The weight of the indicator, is the total number of indicators.

[0062] Step 405: According to the comprehensive weight of each non-periodic test moment, the weight of each indicator is determined using the calculation formula of the indicator weight.

[0063] In another exemplary embodiment of the present application, the process of determining the reliability of each indicator in the above step 106 can be replaced by the following steps 501 to 505.

[0064] Step 501: Using the Monte Carlo method, the data of each indicator in the test data is sampled multiple times to obtain multiple samples.

[0065] Using the Monte Carlo method to No. Indicators Assume that The sample data of samples is ,but The sample data is .

[0066] Step 502: If the sample data of each indicator is not , it is judged as unreliable and recorded as 1; among them, and They are the minimum and maximum values ​​of the data allowed range.

[0067] Step 503: If the sample data of each indicator is If the value is medium, it is judged as reliable and recorded as 0.

[0068] Step 504: Using the formula , determine the reliability of each indicator in each sample; where, For the The indicator is Reliability in samples, For the The first indicator The data of a sample is judged to be unreliable and takes the value 1, or reliable and takes the value 0. is the sample size.

[0069] Step 505: The average reliability of each indicator in all samples is taken as the reliability of each indicator.

[0070] The reliability of each indicator reflects the internal credibility of the evidence.

[0071] In another exemplary embodiment of the present application, the cohesion ratio of the evidential reasoning algorithm in the above step 107 when fusing various indicators in the test data is as follows.

[0072] .

[0073] In the formula, For the The cohesion ratio of the indicators, For the The weight of the indicator, For the The reliability of an indicator.

[0074] In order to distinguish it from weight, the weight generated by integrating the weight of the indicator with the reliability is called the cohesion ratio.

[0075] In another exemplary embodiment of the present application, after obtaining the weight of each indicator and the reliability of each indicator, evidence fusion is performed. The process of evidence fusion can be seen in the following content.

[0076] Consider the evidence of cohesion ratio The basic probability mass distribution of is as follows.

[0077] .

[0078] in Indicates the result assigned to the mth evidence The basic probability mass of . Show evidence Point to results confidence level. Represents a power set. Represents the recognition framework, , Indicates the first, second, ..., possible outcomes. Evidence That is the above indicators.

[0079] evidence The confidence distribution form of can be expressed as follows.

[0080] .

[0081] in, Indicates that the mth piece of evidence points to the proposition , Indicates that the mth piece of evidence points to the identification framework The basic probability mass of .

[0082] According to its basic probability mass generation method, fusion The basic probability mass after the evidence can be expressed as follows.

[0083] .

[0084] .

[0085] in Indicates fusion After the evidence is assigned to the result The basic probability mass of . Representative Fusion The reliability of the evidence. Represents the possible existence of a proposition A collection of . Indicates fusion After the evidence points to the proposition The basic probability mass of Indicates fusion After the evidence points to the power set The basic probability mass of Indicates Evidence points to the proposition The basic probability mass of Indicates fusion After the evidence points to the collection The basic probability mass of Indicates fusion After the evidence points to the collection On the other hand, Represents assignment to the power set The basic probability mass of , represents the global ignorance in the process of evidence fusion.

[0086] Fusion The evidence points to the proposition The probability of , which is calculated by the following formula.

[0087] .

[0088] The final output utility is given by the following function.

[0089] .

[0090] in, Indicates combination The comprehensive performance status of the inertial navigation system after the evidence is obtained, that is, the output utility. Label value representing the output utility.

[0091] The output of the performance evaluation model is the average health level, also known as the utility value. For example, {80% is good, 20% is medium, and 0% is poor}. When {good, medium, poor} are assigned values ​​of {3, 2, 1}, the utility value is 80%×3+20%×2+0×1=2.8.

[0092] In another exemplary embodiment of the present application, during the operation of the inertial navigation system, as the system performance changes, its operating conditions change accordingly. In order to ensure the diagnostic accuracy, the parameters of the performance evaluation model need to be optimized. The parameters of the performance evaluation model are updated through an optimization algorithm. The method may also include: based on constraints, optimizing the parameters of the performance evaluation model in real time; the constraints include the following conditions.

[0093] .

[0094] .

[0095] .

[0096] .

[0097] In the formula, For the The weight of the indicator, For the The reliability of the indicator, is the total number of indicators, is the first reference value, is the lower limit of the reference value during the optimization process. is the upper limit of the reference value during the optimization process. For the L A reference value.

[0098] Mean Square Error (MSE) is an indicator that reflects the degree of difference between the estimated value and the estimated value. and estimated output utility The MSE between can reflect the accuracy of the performance evaluation model. Therefore, the optimization objective function is as follows.

[0099] .

[0100] .

[0101] in, For the moment The estimated output utility of For the moment The actual output utility, For total time.

[0102] This application aims to solve the problems faced in the performance evaluation process of inertial navigation systems, such as the lack of monitoring samples, the difficulty in establishing accurate mathematical models, and the non-periodic characteristics of inertial navigation system test data. The evidence reasoning algorithm implements the design of an inertial navigation system performance evaluation model to obtain the performance status of the inertial navigation system. The main steps of this application can be divided into the following steps: First, the weights of each non-periodic test moment are obtained and fused with the current test moment data to generate new test data. Then, the evidence weight and evidence reliability are obtained based on the historical non-periodic test moment weights to provide parameters for evidence fusion; finally, the ER algorithm is used to fuse the evidence to obtain the performance status of the inertial navigation system.

[0103] The method of this application is experimentally verified below.

[0104] This application uses a certain type of inertial navigation system to carry out experimental research. The gravity acceleration at the experimental site is 9.8015m / s 2 , the inertial navigation system has been in operation for 52 months. The inertial navigation system has been tested 24 times before the current moment, and the time intervals between these 24 tests are not the same. According to expert experience and product description, the three-axis zero-order drift coefficient of the gyroscope (D0x, D0y, D0z) and the three-axis zero-order drift coefficient of the acceleration (K0x, K0y, K0z) are selected as indicators, a total of 6. In order to illustrate the validity of the data, the average data of each indicator during each test is shown as follows Figure 2 shown.

[0105] exist Figure 2 In the time scale of the horizontal axis, unit 1 represents one month. Figure 2 It can be seen that as time goes by, the zero-order drift coefficient of the gyroscope and accelerometer gradually increases. The performance status of the equipment is divided into three levels: {good, medium, and poor}, represented by {G1, G2, and G3} respectively. The performance evaluation results of the inertial navigation system in the previous test process are shown in Figure 3 shown.

[0106] exist Figure 3 In the data, green, yellow and orange correspond to the three performance status levels of good, medium and poor. Taking the 10th group as an example, the probability of the performance status of the inertial navigation system being rated as good is 69.1%, the probability of being rated as medium is 30.9%, and the probability of being rated as poor is 0%. For another example, in the 22nd group of data, the probability of the performance status of the inertial navigation system being rated as good, medium and poor is (0%, 84.8%, 15.2%) respectively. Figure 3 It can be seen that the comprehensive performance status of the inertial navigation system shows a downward trend with the increase of time, which is consistent with the logic of equipment performance degradation.

[0107] Based on the GJB 729-89 standard and expert knowledge, the reference values ​​of various indicators are set as shown in Table 1. The reference values ​​after historical data optimization are shown in Table 2.

[0108] Table 1 Index reference values

[0109]

[0110] Table 2 Index reference values

[0111]

[0112] In Table 1 and Table 2, A1 is the first reference value, and A2 is the second reference value.

[0113] In this experiment, a total of 6000 sets of test data are generated during the test of the three indicators at the current moment. In view of the large amount of data, this application uses a random sampling method to extract 200 sets of data, such as Figure 4 shown.

[0114] The average value of the indicator in the current period is used as the current test value for subsequent calculations. Figure 4 The average values ​​of the data for each indicator can be obtained as shown in Table 3.

[0115] Then, combine the average value of the current indicator data with Figure 2The historical test data in can be used to obtain the bidirectional weights of each non-periodic test moment as follows Figure 5 shown.

[0116] Table 3 Index reference values

[0117]

[0118] Combined with the calculation formula of comprehensive weight, the non-periodic test time weight can be obtained as follows: Figure 6 At the same time, according to the fusion formula of the data of each historical non-periodic test moment and the data of the current test moment and steps 401 to 505, the current new input value and the parameters of the evidence can be calculated as shown in Table 4.

[0119] Table 4 Index reference values

[0120]

[0121] Then, evidence fusion was performed according to the cohesion ratio formula and the formulas involved in the evidence fusion process, and the performance status evaluation results of the inertial navigation system at the current moment were obtained, which were 62.7% for medium and 37.3% for poor.

[0122] In order to verify the stability of the proposed method, this application randomly selected 50 test data for experiments. The results are as follows Figure 7 As shown. Figure 7 It can be seen that the variance of the results of fifty experiments is 0.0386, which is far lower than the reasonable value of stability, indicating that the method of the present application has good stability.

[0123] This application effectively integrates expert knowledge with monitoring information, proposes a two-way weight setting method, and invents a method for fusing non-periodic test indicator data of an inertial navigation system, effectively solving the problem that the performance evaluation accuracy of the inertial navigation system is affected by non-periodic test data. This application effectively improves the performance evaluation accuracy of the inertial navigation system in actual environments, determines the performance status of the inertial navigation system under non-periodic test data, and ensures the safe and reliable operation of the inertial navigation system.

[0124] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0125] The present application uses specific examples to illustrate the principles and implementation methods of the present application. The above examples are only used to help understand the method and core ideas of the present application. At the same time, for those skilled in the art, according to the ideas of the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present application.

Claims

1. A method for evaluating the performance status of an inertial navigation system considering non-periodic data, characterized in that: The inertial navigation system performance status evaluation method considering non-periodic data includes: Establish a performance evaluation model for inertial navigation systems; Determine the lateral weight and longitudinal weight of each non-periodic test moment of the inertial navigation system; all non-periodic test moments include multiple historical non-periodic test moments and the current test moment; Combining the lateral weight and the longitudinal weight of each non-periodic test moment, the comprehensive weight of each non-periodic test moment of the inertial navigation system is obtained; According to the comprehensive weight of each non-periodic test moment, the data of each historical non-periodic test moment and the data of the current test moment are integrated to generate test data; each moment in the test data corresponds to multiple indicators; Determine the weight of each indicator based on the comprehensive weight of each non-periodic test moment; Determine the reliability of each indicator based on the test data; According to the weight of each indicator and the reliability of each indicator, the evidential reasoning algorithm is used to fuse the indicators in the test data, and the performance evaluation model is used to obtain the performance status of the inertial navigation system at the current test time; The calculation formula of the horizontal weight is: ； In the formula, For the The horizontal weight of the non-periodic test moment, is the current test time, is the total number of non-periodic test moments, For the Aperiodic test moments; Determine the longitudinal weight of each non-periodic test moment of the inertial navigation system, including: The performance degradation expression is established as: ; In the formula, For the The non-periodic test time and +1 The performance degradation during the time period between non-periodic test moments, For the The performance utility value of the inertial navigation system corresponding to the non-periodic test moment, For the +1 inertial navigation system performance utility value corresponding to the non-periodic test moment, For the The non-periodic test time and +1 time period between non-periodic test moments, is the total number of non-periodic test moments; According to the performance degradation degree expression, the weight expression of the time period is determined as follows: ; In the formula, For the The non-periodic test time and +The weight of the time period between 1 non-periodic test moments, || is the absolute value symbol; According to the weight expression of the time period, the calculation formula for the longitudinal weight is: ; In the formula, For the The longitudinal weight of the non-periodic test moment, is the weight of the time period between the first non-periodic test moment and the second non-periodic test moment, For the +1 non-periodic test moment and +The weight of the time period between 2 non-periodic test moments; According to the calculation formula of the longitudinal weight, the longitudinal weight of the inertial navigation system at each non-periodic test moment is obtained.

2. The method for evaluating the performance status of an inertial navigation system considering non-periodic data according to claim 1, characterized in that: The performance evaluation model is: ； In the formula, The performance evaluation model indicators; For the Considered in the index Data of non-periodic test moments, There are multiple performance states that may occur in the inertial navigation system. In order The corresponding possibility, For the The weight of the indicator, For the The reliability of the indicator, for A non-periodic test moment.

3. The method for evaluating the performance status of an inertial navigation system considering non-periodic data according to claim 1, characterized in that: The calculation formula of the comprehensive weight is: ； In the formula, For the inertial navigation system The comprehensive weight of the non-periodic test moments, For the The horizontal weight of the non-periodic test moment, For the The vertical weight of each non-periodic test moment.

4. The method for evaluating the performance status of an inertial navigation system considering non-periodic data according to claim 1, characterized in that: According to the comprehensive weight of each non-periodic test moment, the data of each historical non-periodic test moment and the data of the current test moment are integrated to generate test data, including: According to the comprehensive weight of each non-periodic test moment, according to the formula , fuse the data of each historical non-periodic test time with the data of the current test time to obtain the fused non-periodic test data; where, is the fused non-periodic test data, is the data at the current test time, For the inertial navigation system The comprehensive weight of the non-periodic test moments, For the Data of non-periodic test moments; When each non-periodic test moment corresponds to When there are 1 index, the fused non-periodic test data is expressed as: ;in, For input, is the input value of the first indicator in the fused non-periodic test data, is the first The input value of the indicator, is the first The input value of the indicator, is the confidence level of the first indicator, For the The confidence level of an indicator, For the The confidence level of each indicator; Convert each item in the fused non-periodic test data into a confidence distribution based on the indicator reference value: ; In the formula, For the The first indicator Results for point to No. Reference values, for point to No. +1 reference value, is the number of results; All confidence distribution forms based on the index reference values ​​constitute the test data.

5. The method for evaluating the performance status of an inertial navigation system considering non-periodic data according to claim 1, characterized in that: According to the comprehensive weight of each non-periodic test moment, the weight of each indicator is determined, including: The calculation formula for the first coefficient of variation is established as: ; In the formula, For the The first coefficient of variation of an indicator at the current test moment; For the The average value of the indicators, For the The root mean square error of each indicator; is the current test time; The calculation formula for establishing the second coefficient of variation is: ; In the formula, is the second coefficient of variation at the current test moment, For the inertial navigation system The comprehensive weight of the non-periodic test moments, It is the comprehensive weighted average value of the inertial navigation system at all historical non-periodic test moments; According to the calculation formula of the first coefficient of variation and the calculation formula of the second coefficient of variation, the expression of the comprehensive coefficient of variation is determined as follows: ; In the formula, For the The comprehensive coefficient of variation of the indicators at the current test time, is the dynamic adjustment coefficient; The calculation formula for establishing indicator weights is: ; In the formula, For the The weight of the indicator, is the total number of indicators; According to the comprehensive weight of each non-periodic test moment, the weight of each indicator is determined using the calculation formula of the indicator weight.

6. The method for evaluating the performance status of an inertial navigation system considering non-periodic data according to claim 1, characterized in that: According to the test data, the reliability of each indicator is determined, including: Using the Monte Carlo method, the data of each indicator in the test data is sampled multiple times to obtain multiple samples; If the sample data of each indicator is not , it is judged as unreliable and recorded as 1; among them, and They are the minimum and maximum values ​​of the data allowed range respectively; If the sample data of each indicator is If the value is medium, it is judged as reliable and recorded as 0; Using the formula , determine the reliability of each indicator in each sample; where, For the The indicator is Reliability in samples, For the The first indicator The data of a sample is judged to be unreliable and takes the value 1, or reliable and takes the value 0. is the sample size; The average reliability of each indicator in all samples is taken as the reliability of each indicator.

7. The method for evaluating the performance status of an inertial navigation system considering non-periodic data according to claim 1, characterized in that: The cohesion ratio of the evidential reasoning algorithm when integrating the indicators in the test data is: ； In the formula, For the The cohesion ratio of the indicators, For the The weight of the indicator, For the The reliability of an indicator.

8. The method for evaluating the performance status of an inertial navigation system considering non-periodic data according to claim 1, characterized in that: The method for evaluating the performance status of an inertial navigation system considering non-periodic data further includes: Optimize the parameters of the performance evaluation model in real time based on constraints and optimization objective functions; The constraints include: ； ； ； ； In the formula, For the The weight of the indicator, For the The reliability of the indicator, is the total number of indicators, is the first reference value of the indicator. is the lower limit of the reference value during the optimization process. The upper limit of the reference value in the optimization process. For the L Reference values; The optimization objective function is: ； ； In the formula, is the mean square error, For the The first indicator Results For the moment The estimated output utility of For the moment The actual output utility, For total time.

Images