Aircraft hydraulic oil leakage fault diagnosis and prediction method based on a two-step correction strategy

Abstract

The embodiment of this specification provides an aircraft hydraulic oil leakage fault diagnosis and prediction method based on a two-step correction strategy. The aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy includes: obtaining at least two monitoring parameters of the hydraulic system of the target device within a historical time interval, where the at least two monitoring parameters include a hydraulic oil quantity parameter; correcting the parameter value of the hydraulic oil quantity parameter to generate a characteristic value of the hydraulic oil quantity parameter; in the case of determining that the characteristic value belongs to the parameter value interval corresponding to the pre-generated hydraulic oil quantity parameter, updating the parameter value interval to generate a target parameter value interval; inputting the characteristic value into a pre-trained fault detection model for processing to generate a predicted value of the hydraulic oil quantity of the hydraulic system at the target moment; and determining a fault detection result of the hydraulic system at the target moment based on the predicted value of the hydraulic oil quantity and the target parameter value interval.

Description

Technical Field

[0001] The embodiments of this specification relate to the technical field of hydraulic systems, and particularly to an aircraft hydraulic oil leakage fault diagnosis and prediction method based on a two-step correction strategy. Background Art

[0002] A hydraulic system refers to a complete set of devices that uses hydraulic oil as the working medium, relies on the hydraulic oil to transmit pressure, and drives the actuator to complete specific operating actions. The basic components of a hydraulic system include a hydraulic oil pump, control valves, actuators, etc. The hydraulic system has many excellent properties, such as smooth operation, high reliability, large power-to-mass ratio, and convenient implementation of automatic control. Therefore, since the 1940s, hydraulic systems have been widely used in various equipment. In the 1960s, the rapid development of electrical and electronic technologies and automatic control theories led to the hydraulic technology becoming one of the popular research directions in the field of mechatronics, and thus promoted the rapid development of hydraulic system technologies. Hydraulic systems have gradually become an important and indispensable part of many modern devices.

[0003] To improve the reliability of hydraulic systems, diagnosing and predicting hydraulic system leakage faults is an important method. However, currently, the diagnostic and prediction methods for hydraulic system leakage faults in many devices still remain in the stage of manual discrimination. For example, visually inspecting, regularly inspecting, etc. to observe whether there is oil leakage in the hydraulic system; or directly observing the change trend of the hydraulic system oil quantity data within a certain period of time to manually judge whether there is oil leakage in the hydraulic system. This manual discrimination method relies on experience and has a low diagnostic frequency, and cannot achieve intelligent diagnosis and active prevention. Therefore, how to achieve accurate intelligent diagnosis and prediction of hydraulic oil leakage faults is an urgent problem to be solved in the current field of hydraulic system fault diagnosis. Summary of the Invention

[0004] In view of this, the embodiments of this specification provide an aircraft hydraulic oil leakage fault diagnosis and prediction method based on a two-step correction strategy. One or more embodiments of this specification also relate to an aircraft hydraulic oil leakage fault diagnosis and prediction device based on a two-step correction strategy, a computing device, a computer-readable storage medium, and a computer program to solve the technical defects existing in the prior art.

[0005] According to the first aspect of the embodiments of this specification, there is provided an aircraft hydraulic oil leakage fault diagnosis and prediction method based on a two-step correction strategy, including:

[0006] Obtaining at least two monitoring parameters of the hydraulic system of the target device within a historical time interval, where the at least two monitoring parameters include hydraulic oil quantity parameters;

[0007] Modify the parameter value of the hydraulic oil quantity parameter to generate the characteristic value of the hydraulic oil quantity parameter;

[0008] When it is determined that the characteristic value belongs to the parameter value interval corresponding to the pre-generated hydraulic oil quantity parameter, update the parameter value interval to generate the target parameter value interval;

[0009] Input the characteristic value into a pre-trained fault detection model for processing to generate the predicted value of the hydraulic oil quantity of the hydraulic system at the target moment;

[0010] Based on the predicted value of the hydraulic oil quantity and the target parameter value interval, determine the fault detection result of the hydraulic system at the target moment.

[0011] Optionally, the modification of the parameter value of the hydraulic oil quantity parameter includes:

[0012] Determine the first parameter value of the hydraulic oil quantity parameter of the hydraulic system at the current startup moment, and determine the target shutdown duration of the hydraulic system according to the current startup moment and the previous shutdown moment before the current startup moment;

[0013] Based on the target shutdown duration and the correlation between the shutdown duration and the hydraulic oil increase amount, modify the first parameter value to generate the second parameter value.

[0014] Optionally, the at least two detection parameters further include the historical shutdown duration and the corresponding hydraulic oil increase amount during the historical duration;

[0015] Correspondingly, the modification of the first parameter value based on the target shutdown duration and the correlation between the shutdown duration and the hydraulic oil increase amount includes:

[0016] Determine the reference shutdown duration of the hydraulic system according to multiple historical shutdown durations within the historical time interval;

[0017] Establish an empirical function between the shutdown duration and the hydraulic oil increase amount according to the corresponding relationship between the multiple historical shutdown durations and the multiple hydraulic oil increase amounts;

[0018] Based on the target shutdown duration, the reference shutdown duration, and the empirical function, modify the first parameter value.

[0019] Optionally, the at least two detection parameters further include the historical hydraulic oil sudden increase amount of the hydraulic system when the target device releases the parking brake;

[0020] Correspondingly, the modification of the parameter value of the hydraulic oil quantity parameter includes:

[0021] Determine whether the target device releases the parking brake before the engine of the target device starts;

[0022] If not, determine the target hydraulic oil sudden increase of the hydraulic system when the brake is released according to multiple historical hydraulic oil sudden increases within the historical time interval;

[0023] Correct the second parameter value based on the target hydraulic oil sudden increase.

[0024] Optionally, the aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy further includes:

[0025] Determine multiple historical characteristic values of the hydraulic oil quantity parameter within the historical time interval;

[0026] Calculate the mean value and standard deviation corresponding to the multiple historical characteristic values;

[0027] Based on the mean value and the standard deviation, and using the three-sigma principle, construct a parameter value interval corresponding to the hydraulic oil quantity parameter.

[0028] Optionally, updating the parameter value interval to generate a target parameter value interval includes:

[0029] Use the upper and lower limit values of the parameter value interval as prior information, use the characteristic value and at least two historical characteristic values among the multiple historical characteristic values as posterior information, and use Bayesian estimation to update the upper and lower limit values of the parameter value interval to generate a target parameter value interval.

[0030] Optionally, the fault detection model includes an encoder, a decoder, and a prediction network. The encoder and the decoder are each composed of two layers of long short-term memory networks, and the prediction network is composed of four layers of fully connected layers;

[0031] Correspondingly, inputting the characteristic value into a pre-trained fault detection model for processing includes:

[0032] Determine multiple historical characteristic values of the hydraulic oil quantity parameter within the historical time interval;

[0033] Use the characteristic value and at least two historical characteristic values among the multiple historical characteristic values as input data and input them into the encoder, and perform encoding processing on the input data through the encoder to generate corresponding embedding vectors;

[0034] Input the embedding vectors into the prediction network for processing.

[0035] Optionally, the aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy further includes:

[0036] Perform sliding segmentation on the multiple historical feature values to generate sample data and sample labels;

[0037] Input the sample data into the encoder of the to-be-trained fault detection model for encoding processing to generate a first embedding vector;

[0038] Input the first embedding vector into the prediction network of the to-be-trained fault detection model for processing to generate corresponding prediction results;

[0039] Calculate a first loss value of the to-be-trained fault detection model based on the sample labels and the prediction results;

[0040] Calculate a first gradient value corresponding to the network parameters of the prediction network according to the chain rule of differentiation in combination with the first loss value;

[0041] Update the network parameters according to the gradient descent method in combination with the first gradient value;

[0042] When it is determined that the loss function of the to-be-trained fault detection model converges according to the first loss value, stop training to obtain the fault detection model.

[0043] Optionally, the aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy further includes:

[0044] Input the sample data into the encoder of the initial fault detection model for encoding processing to generate a second embedding vector;

[0045] Input the second embedding vector into the decoder of the initial fault detection model for decoding processing to generate reconstructed sample data corresponding to the sample data;

[0046] Calculate a second loss value of the initial fault detection model based on the sample data and the reconstructed sample data;

[0047] Calculate second gradient values corresponding to the network parameters of the encoder and the decoder according to the chain rule of differentiation in combination with the second loss value;

[0048] Update the network parameters of the encoder and the decoder according to the gradient descent method in combination with the second gradient values;

[0049] When it is determined that the loss function of the initial fault detection model converges according to the second loss value, stop training to obtain the to-be-trained fault detection model.

[0050] According to the second aspect of the embodiments of the present specification, there is provided an aircraft hydraulic oil leakage fault diagnosis and prediction device based on a two-step correction strategy, including:

[0051] An acquisition module, configured to acquire at least two monitoring parameters of a hydraulic system of a target device within a historical time interval, where the at least two monitoring parameters include a hydraulic oil quantity parameter;

[0052] A correction module, configured to correct the parameter value of the hydraulic oil quantity parameter to generate a characteristic value of the hydraulic oil quantity parameter;

[0053] An update module, configured to update the parameter value interval to generate a target parameter value interval when it is determined that the characteristic value belongs to a pre-generated parameter value interval corresponding to the hydraulic oil quantity parameter;

[0054] A processing module, configured to input the characteristic value into a pre-trained fault detection model for processing to generate a predicted value of the hydraulic oil quantity of the hydraulic system at a target moment;

[0055] A determination module, configured to determine a fault detection result of the hydraulic system at the target moment based on the predicted value of the hydraulic oil quantity and the target parameter value interval.

[0056] According to a third aspect of the embodiments of the present specification, a computing device is provided, including:

[0057] A memory and a processor;

[0058] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the steps of any one of the above-mentioned aircraft hydraulic oil leakage fault diagnosis and prediction methods based on a two-step correction strategy.

[0059] According to a fourth aspect of the embodiments of the present specification, a computer-readable storage medium is provided, which stores computer-executable instructions, and when the instructions are executed by a processor, the steps of any one of the above-mentioned aircraft hydraulic oil leakage fault diagnosis and prediction methods based on a two-step correction strategy are implemented.

[0060] According to a fifth aspect of the embodiments of the present specification, a computer program is provided, where when the computer program is executed on a computer, the computer is made to execute the steps of the above-mentioned aircraft hydraulic oil leakage fault diagnosis and prediction method based on a two-step correction strategy.

[0061] The aircraft hydraulic oil leakage fault diagnosis and prediction method based on a two-step correction strategy provided by the embodiments of this specification obtains at least two monitoring parameters of the hydraulic system of the target device within a historical time interval, where the at least two monitoring parameters include a hydraulic oil quantity parameter; corrects the parameter value of the hydraulic oil quantity parameter to generate a characteristic value of the hydraulic oil quantity parameter; updates the parameter value interval to generate a target parameter value interval when determining that the characteristic value belongs to the pre-generated parameter value interval corresponding to the hydraulic oil quantity parameter; inputs the characteristic value into a pre-trained fault detection model for processing to generate a predicted value of the hydraulic oil quantity of the hydraulic system at the target moment; and determines the fault detection result of the hydraulic system at the target moment based on the predicted value of the hydraulic oil quantity and the target parameter value interval. By correcting the parameter value of the hydraulic oil quantity parameter, not only the oil quantity in the hydraulic oil tank is considered, but also the hydraulic oil quantity in the hydraulic pipeline is considered, making the corrected hydraulic oil quantity closer to the actual oil quantity, with better authenticity and predictability; in addition, by updating the parameter value interval, it is beneficial to ensure the long-term effectiveness of the upper and lower threshold values of this interval and can effectively reduce the false alarm rate; furthermore, based on the fault detection model, the hydraulic oil quantity is predicted, thereby realizing the fault prediction of the hydraulic system, which is beneficial to improving the accuracy and stability of the prediction result. Description of the Drawings

[0062] Figure 1 is a flowchart of a method for diagnosing and predicting aircraft hydraulic oil leakage faults based on a two-step correction strategy provided by an embodiment of this specification;

[0063] Figure 2 is a schematic structural diagram of a fault detection model provided by an embodiment of this specification;

[0064] Figure 3 is a flowchart of the processing procedure of a method for diagnosing and predicting aircraft hydraulic oil leakage faults based on a two-step correction strategy provided by an embodiment of this specification;

[0065] Figure 4 is a flowchart of the process for correcting the parameter value of a hydraulic oil quantity parameter provided by an embodiment of this specification;

[0066] Figure 5 is a flowchart of the process for updating a parameter value interval provided by an embodiment of this specification;

[0067] Figure 6 is a flowchart of a fault detection process provided by an embodiment of this specification;

[0068] Figure 7 is a schematic structural diagram of a device for diagnosing and predicting aircraft hydraulic oil leakage faults based on a two-step correction strategy provided by an embodiment of this specification;

[0069] Figure 8 It is a structural block diagram of a computing device provided by an embodiment of this specification. Detailed implementation manners

[0070] In the following description, numerous specific details are set forth in order to provide a thorough understanding of this specification. However, this specification can be implemented in many other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the connotation of this specification. Therefore, this specification is not limited by the specific implementations disclosed below.

[0071] The terms used in one or more embodiments of this specification are for the purpose of describing specific embodiments only and are not intended to limit one or more embodiments of this specification. The singular forms "a", "the", and "said" used in one or more embodiments of this specification and the appended claims are also intended to include the plural forms unless the context clearly dictates otherwise. It should also be understood that the term "and / or" used in one or more embodiments of this specification refers to and encompasses any and all possible combinations of one or more of the associated listed items.

[0072] It should be understood that although the terms first, second, etc. may be used in one or more embodiments of this specification to describe various information, such information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of one or more embodiments of this specification, the first may also be referred to as the second, and similarly, the second may also be referred to as the first. Depending on the context, the word "if" as used herein can be interpreted as "when" or "while" or "in response to determining".

[0073] First, the noun terms related to one or more embodiments of this specification are explained.

[0074] Currently, most aircraft use electro-hydraulic servo systems as the actuation systems of the aircraft, and the hydraulic actuation systems of the aircraft undertake the opening, closing, and other complex control functions of the moving parts on the aircraft. For example, the retraction and control of important components such as aircraft wheels, control surfaces, speed brakes, flaps, and engine reversers are all undertaken by the hydraulic system. To ensure the reliability of the hydraulic system operation, generally two or more sets of independent hydraulic systems are equipped on the aircraft. One of the hydraulic systems undertakes the main control responsibility and is called the common hydraulic system; the remaining hydraulic systems undertake secondary control responsibilities or are used as backups and are called boost hydraulic systems (or control hydraulic systems). The common hydraulic system is generally used for the retraction and extension of components such as landing gears, speed brakes, and flaps, front wheel steering control, driving fuel pump motors, etc. The boost hydraulic system is only used to drive boosters and damping servo actuators of the above-mentioned flight control systems.

[0075] The health status of an aircraft hydraulic system directly affects the flight safety of the aircraft. According to statistics, among all mechanical failures of an aircraft, hydraulic system failures account for 30%. Once a failure occurs in the aircraft hydraulic system, on the one hand, it will bring great instability factors to the flight process and cause the failure of the flight mission; on the other hand, if the emergency equipment cannot effectively cope with the environmental conditions when the hydraulic system fails, it may cause serious consequences of plane crash and death. Therefore, it is crucial to diagnose and predict aircraft hydraulic system failures in a timely and accurate manner.

[0076] Among the hydraulic system failure modes, the leakage failures of hydraulic oil, including running, leaking, dripping, etc., account for the vast majority. Therefore, the diagnosis and prediction of hydraulic system leakage failures are important methods to improve the reliability of the hydraulic system.

[0077] In this specification, a method for diagnosing and predicting aircraft hydraulic oil leakage failures based on a two-step correction strategy is provided. This specification also relates to a device for diagnosing and predicting aircraft hydraulic oil leakage failures based on a two-step correction strategy, a computing device, a computer-readable storage medium, and a computer program, which will be described in detail one by one in the following embodiments.

[0078] Figure 1 The flowchart of a method for diagnosing and predicting aircraft hydraulic oil leakage failures based on a two-step correction strategy provided according to an embodiment of this specification is shown, which specifically includes the following steps.

[0079] Step 102: Obtain at least two monitoring parameters of the hydraulic system of the target device within a historical time interval, where the at least two monitoring parameters include a hydraulic oil quantity parameter.

[0080] In the embodiments of this specification, the target device includes, but is not limited to, machine tools and metal processing equipment, construction machinery and engineering equipment, metallurgical and mining equipment, aerospace and military equipment, automobiles and transportation vehicles, light industrial equipment, and heavy industrial equipment, etc. However, in the embodiments of this specification, only the target device being an aircraft is taken as an example for illustration. The process of diagnosing and predicting aircraft hydraulic oil leakage failures based on a two-step correction strategy for other types of target devices is the same as the fault detection process of the aircraft hydraulic system, which will not be elaborated here.

[0081] Step 104: Correct the parameter value of the hydraulic oil quantity parameter to generate a characteristic value of the hydraulic oil quantity parameter.

[0082] In an optional implementation manner, the correcting the parameter value of the hydraulic oil quantity parameter includes:

[0083] Determine the first parameter value of the hydraulic oil quantity parameter of the hydraulic system at the current startup moment, and determine the target shutdown duration of the hydraulic system according to the current startup moment and the previous shutdown moment before the current startup moment;

[0084] Based on the target shutdown duration and the correlation between the shutdown duration and the hydraulic oil increase amount, correct the first parameter value to generate a second parameter value.

[0085] Further, the at least two detection parameters further include the historical shutdown duration and the hydraulic oil increase amount corresponding to the historical duration;

[0086] Correspondingly, the correcting the first parameter value based on the target shutdown duration and the correlation between the shutdown duration and the hydraulic oil increase amount includes:

[0087] Determine the reference shutdown duration of the hydraulic system according to multiple historical shutdown durations within the historical time interval;

[0088] Establish an empirical function between the shutdown duration and the hydraulic oil increase amount according to the corresponding relationship between the multiple historical shutdown durations and the multiple hydraulic oil increase amounts;

[0089] Based on the target shutdown duration, the reference shutdown duration, and the empirical function, correct the first parameter value.

[0090] Even further, the at least two detection parameters further include the historical hydraulic oil sudden increase amount of the hydraulic system when the target device releases the parking brake;

[0091] Correspondingly, the correcting the parameter value of the hydraulic oil quantity parameter includes:

[0092] Judge whether the target device releases the parking brake before the engine of the target device starts;

[0093] If not, determine the target hydraulic oil sudden increase amount of the hydraulic system when the brake is released according to multiple historical hydraulic oil sudden increase amounts within the historical time interval;

[0094] Based on the target hydraulic oil sudden increase amount, correct the second parameter value.

[0095] Specifically, correcting the parameter value of the hydraulic oil quantity parameter is to correct the hydraulic oil tank quantity of the hydraulic system at the engine startup moment of this time. Before correcting it, it is necessary to determine the parameters of the extrapolation correction and the brake release correction algorithm according to historical data. In the embodiments of this specification, at least two monitoring parameters obtained within the historical time interval include the hydraulic oil tank quantity of the hydraulic system at the engine startup moment and other historical data.

[0096] The parameters of the extrapolation correction algorithm include the reference shutdown duration , and the empirical function of the increase in hydraulic oil volume varying with the shutdown duration. Here, the shutdown duration refers to the difference between the engine startup time and the previous engine shutdown time. Calculate the historical shutdown durations in the historical data , and take the mode of the historical shutdown durations as the reference shutdown duration . The increase in hydraulic oil volume during the shutdown duration refers to the difference between the hydraulic oil tank volume corresponding to the hydraulic system at the engine startup time and the hydraulic oil tank volume corresponding to the hydraulic system at the previous engine shutdown time. Calculate the increase in hydraulic oil volume corresponding to the historical shutdown times in the historical data . As the hydraulic oil in the hydraulic pipeline continuously flows into the oil tank, the increase rate of the oil volume in the tank follows the law of decreasing from fast to slow. Therefore, the relationship between the increase in hydraulic oil volume and the historical shutdown time approximately follows a logarithmic function . Transform this formula into . According to the historical shutdown duration and the corresponding increase in hydraulic oil volume , the empirical function of the increase in hydraulic oil volume varying with the historical shutdown time can be obtained by using the least squares method as follows:

[0097]

[0098] The parameters of the brake release correction algorithm include the target sudden increase in hydraulic oil volume of the hydraulic system when the brake is released . Since the parking brake may be released before the engine startup for some aircraft, the release of the parking brake will cause the hydraulic oil sealed in the brake to flow back to the hydraulic oil tank, resulting in a small increase in the hydraulic oil volume. For aircraft that do not release the parking brake before the engine startup, in order to make the extracted hydraulic oil volume closer to the actual hydraulic oil volume, the impact of this sudden increase needs to be considered. Statistically analyze the magnitudes of multiple historical sudden increases in hydraulic oil volume in the historical data, and take the average of the multiple historical sudden increases in hydraulic oil volume as the target sudden increase in hydraulic oil volume of the hydraulic system when the brake is released .

[0099] When correcting the parameter values of the hydraulic oil volume parameters, first adaptively extract the first parameter value of the hydraulic oil volume parameters of the hydraulic system at the current startup time and the target shutdown duration . Then, take the first parameter value as the hydraulic oil volume reference value , and perform extrapolation correction on the hydraulic oil volume reference value according to the target shutdown duration and the reference shutdown duration :

[0100]

[0101] ​​The corrected second parameter value is .

[0102] Then, it is determined whether the acquired data contains relevant data on "brake release action before engine start of the aircraft". If it exists, no brake release correction is performed, and is used as the corrected eigenvalue; if it does not exist, brake release correction is performed:

[0103]

[0104] The corrected eigenvalue is .

[0105] In the embodiment of this specification, the input hydraulic oil quantity data is corrected in two steps, namely extrapolation correction and brake release correction, to form a hydraulic oil quantity eigenvalue. Specifically, first, the hydraulic oil tank quantity data at the moment before engine start is extracted as the reference value. Then, the aircraft parking duration is extracted, and the hydraulic oil quantity extrapolation correction value is determined according to the actual parking duration, the reference parking duration, and the empirical formula. The extrapolation correction is realized by adding the extrapolation correction value to the reference hydraulic oil quantity. Finally, it is determined whether the parking brake is released before the engine starts. If the parking brake has been released, no brake release correction is performed; if the parking brake has not been released, the brake release correction is realized by adding the correction value to the hydraulically oil quantity after extrapolation correction. Finally, the hydraulically oil quantity after the two corrections is used as the hydraulic oil quantity eigenvalue. By performing extrapolation correction and brake release correction in this way, not only the hydraulic oil quantity in the hydraulic oil tank is considered, but also the hydraulic oil quantity in the hydraulic pipeline is considered, making the corrected hydraulic oil quantity closer to the actual oil quantity, with better authenticity and predictability.

[0106] Step 106: In the case where it is determined that the eigenvalue belongs to the parameter value interval corresponding to the pre-generated hydraulic oil quantity parameter, update the parameter value interval to generate a target parameter value interval.

[0107] In an alternative embodiment, the aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy further includes:

[0108] Determine multiple historical eigenvalues of the hydraulic oil quantity parameter within the historical time interval;

[0109] Calculate the mean and standard deviation corresponding to the multiple historical eigenvalues;

[0110] Based on the mean and the standard deviation, and using the three-sigma principle, construct the parameter value interval corresponding to the hydraulic oil quantity parameter.

[0111] Furthermore, the updating the parameter value interval to generate a target parameter value interval includes:

[0112] Taking the upper and lower limit values of the parameter value interval as prior information, taking the eigenvalue and at least two historical eigenvalue among the multiple historical eigenvalues as posterior information, and using Bayesian estimation to update the upper and lower limit values of the parameter value interval to generate a target parameter value interval.

[0113] Specifically, before updating the parameter value interval, the upper and lower limit thresholds of the eigenvalue of the hydraulic oil quantity parameter can be determined first according to multiple historical eigenvalue data of the hydraulic oil quantity parameter within the historical time interval. Determine the upper and lower limit thresholds of the eigenvalue of the hydraulic oil quantity parameter. Assume that the eigenvalue data of the hydraulic oil quantity parameter follows a normal distribution, and use the three-sigma principle to construct the upper and lower limit thresholds:

[0114]

[0115] The parameter value interval composed of the upper and lower limit thresholds is .

[0116] In the case of correcting and generating the second parameter value, it can be first judged whether the second parameter value is within the parameter value interval; if not, it is determined that the second parameter value is abnormal, that is, a hydraulic oil leakage fault has occurred. In this case, a fault warning needs to be issued; if so, it is determined that the second parameter value is normal and no hydraulic oil leakage fault has occurred. Therefore, the second parameter value can be used to update the aforementioned upper and lower limit thresholds.

[0117] During the threshold update process, in order to simplify the calculation, the latest m - 1 historical eigenvalues and the second parameter value are used to update the threshold. Taking the existing threshold as prior information and the aforementioned m samples as posterior information, use Bayesian estimation to update the upper and lower limit thresholds:

[0118]

[0119] Among them, is the probability distribution, are the updated mean, standard deviation, upper limit threshold, and lower limit threshold respectively.

[0120] The target parameter value interval generated after the update is .

[0121] In the embodiments of this specification, threshold construction and threshold update are performed according to the hydraulic oil quantity characteristic values. First, based on the characteristic values extracted from historical data, the upper and lower threshold values of the characteristic values are constructed using the Bayesian estimation method and the three-sigma principle. Then, it is determined whether the newly input oil quantity characteristic value is within the upper and lower threshold values. If it is within the threshold values, it is determined that the data is normal, and the threshold values are updated according to this data; if it is not within the threshold values, it is determined that the data is abnormal, and the threshold values are not updated. The adaptive threshold constructed through historical data can overcome the uncertainty and subjectivity of traditional empirical diagnosis, and the continuous threshold update ensures the long-term effectiveness of the adaptive threshold, which can effectively reduce the false alarm rate.

[0122] Step 108: Input the characteristic value into a pre-trained fault detection model for processing to generate a predicted value of the hydraulic oil quantity of the hydraulic system at the target time.

[0123] In an alternative embodiment, the fault detection model includes an encoder, a decoder, and a prediction network. The encoder and the decoder are each composed of two layers of long short-term memory networks, and the prediction network is composed of four layers of fully connected layers;

[0124] Correspondingly, the inputting the characteristic value into a pre-trained fault detection model for processing includes:

[0125] Determine multiple historical characteristic values of the hydraulic oil quantity parameters within the historical time interval;

[0126] Input the characteristic value and at least two historical characteristic values among the multiple historical characteristic values as input data into the encoder, and perform encoding processing on the input data through the encoder to generate corresponding embedding vectors;

[0127] Input the embedding vectors into the prediction network for processing.

[0128] In another alternative embodiment, the aircraft hydraulic oil leakage fault diagnosis and prediction method based on a two-step correction strategy further includes:

[0129] Input the sample data into the encoder of the initial fault detection model for encoding processing to generate a second embedding vector;

[0130] Input the second embedding vector into the decoder of the initial fault detection model for decoding processing to generate a reconstructed sample data corresponding to the sample data;

[0131] Calculate a second loss value of the initial fault detection model based on the sample data and the reconstructed sample data;

[0132] Calculate a second gradient value corresponding to the network parameters of the encoder and the decoder according to the chain rule of differentiation and in combination with the second loss value;

[0133] Update the network parameters of the encoder and the decoder according to the gradient descent method and in combination with the second gradient value;

[0134] When it is determined that the loss function of the initial fault detection model converges according to the second loss value, stop training to obtain the to-be-trained fault detection model.

[0135] Furthermore, the aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy further includes:

[0136] Perform sliding segmentation on the multiple historical eigenvalue to generate sample data and sample labels;

[0137] Input the sample data into the encoder of the to-be-trained fault detection model for encoding processing to generate a first embedding vector;

[0138] Input the first embedding vector into the prediction network of the to-be-trained fault detection model for processing to generate corresponding prediction results;

[0139] Calculate a first loss value of the to-be-trained fault detection model based on the sample labels and the prediction results;

[0140] Calculate a first gradient value corresponding to the network parameters of the prediction network according to the chain rule of differentiation and in combination with the first loss value;

[0141] Update the network parameters according to the gradient descent method and in combination with the first gradient value;

[0142] When it is determined that the loss function of the to-be-trained fault detection model converges according to the first loss value, stop training to obtain the fault detection model.

[0143] The structural schematic diagram of a fault detection model provided by an embodiment of this specification is as Figure 2 shown. The fault detection model is composed of three parts: an encoder, a decoder, and a prediction network. The encoder and the decoder are each composed of two layers of long short-term memory networks, and the prediction network is composed of four layers of fully connected layers.

[0144] The embodiment of this specification trains and generates a fault detection model through the following steps:

[0145] 1) Determine multiple historical eigenvalue of the hydraulic oil quantity parameter within the historical time interval , and perform sliding segmentation on it to obtain sample data and sample labels Among them, i is the sample serial number, tf is the dimension of the input vector, and tp is the dimension of the output vector. The physical meanings of tf and tp are to predict the future tp eigenvalue with tf historical eigenvalues.

[0146] 2) First, pre-train the encoder and the decoder. During the pre-training process, input the sample data into the encoder first to obtain the embedding vector :

[0147]

[0148] Among them, represents the encoder module.

[0149] Then input the embedding vector into the decoder to restore the sample data to the reconstructed sample data :

[0150]

[0151] Among them, represents the decoder module.

[0152] Then, take the mean absolute error between the sample data and the reconstructed sample data as the loss function :

[0153]

[0154] Calculate the gradient of the model parameters according to the chain rule of differentiation :

[0155]

[0156] In the formula, are the parameters of the encoder and the decoder, including the weight W, the bias B, and other parameters.

[0157] Update the parameters of the encoder and the decoder to according to the gradient descent method:

[0158]

[0159] Among them, represents the learning rate, represents the gradient of with respect to.

[0160] Judge whether the loss function converges. If the loss function has converged, end the training; otherwise, use the updated parameters to replace the current parameters , and repeat the above training process.

[0161] 3) Train the prediction network. During the training process, the encoder parameters are first frozen, and then the sample data is Input into the encoder to get the embedding vector :

[0162]

[0163] In the formula, Represents an encoder module.

[0164] Then embed the vector Input into the prediction network to get the prediction result :

[0165]

[0166] In the formula, Represents the prediction network.

[0167] Then the sample labels And the prediction results The mean absolute error between :

[0168]

[0169] Calculate the gradient corresponding to the network parameters of the prediction network according to the chain rule :

[0170]

[0171] In the formula, To predict the network parameters of the network, including weights , Bias and other parameters.

[0172] According to the gradient descent method, the network parameters of the prediction network are updated as :

[0173]

[0174] in, represents the learning rate, Express gradient.

[0175] Determine whether the loss function has converged. If the loss function has converged, end the training; otherwise, use the updated parameters Replace the current parameters , and repeat the above training process.

[0176] 4) Conduct the hydraulic oil leakage fault prediction process. In this process, the input of the fault detection model is the corrected eigenvalue . First, select the latest tf-1 historical eigenvalues from multiple historical eigenvalues of the hydraulic oil quantity parameter within the historical time interval , and combine it with to form the input data of the fault detection model . Then, input the input data into the encoder to obtain the embedding vector :

[0177]

[0178] In the formula, represents the encoder module.

[0179] Input the embedding vector into the prediction network to obtain the prediction result, that is, the predicted value of the hydraulic oil quantity :

[0180]

[0181] In the formula, represents the prediction network.

[0182] In the embodiment of this specification, the fault detection model is used to predict the input hydraulic oil quantity. First, according to the eigenvalues extracted from historical data, the parameters of the fault detection model are optimized by the gradient descent method to make its training loss smaller. During use, the trained fault detection model is used to predict the hydraulic oil quantity, which has better accuracy and stability compared with traditional prediction methods and can accurately give an early warning of hydraulic oil leakage faults.

[0183] Step 110: Based on the predicted value of the hydraulic oil quantity and the target parameter value interval, determine the fault detection result of the hydraulic system at the target moment.

[0184] After obtaining the prediction result, that is, the predicted value of the hydraulic oil quantity, the fault detection result of the hydraulic system at the target moment can be determined based on this predicted value of the hydraulic oil quantity and the target parameter value interval.

[0185] Specifically, it can be judged whether the predicted value of the hydraulic oil quantity is within the target parameter value interval. If so, it is considered that the probability of a hydraulic oil leakage fault occurring within the future tp moment is relatively low; if not, it is considered that the probability of a hydraulic oil leakage fault occurring within the future tp moment is relatively high, and a fault warning needs to be issued to provide a reference for maintenance work.

[0186] The embodiments of this specification use a two-step correction strategy to obtain corrected hydraulic oil quantity data from the original hydraulic oil tank quantity data. The corrected data has a more realistic representativeness and better predictability. By using the Bayesian method to construct a threshold interval for the corrected hydraulic oil quantity, the fault diagnosis based on the threshold interval overcomes the dependence on professional experience in the manual discrimination method, and can significantly improve the efficiency and accuracy of hydraulic oil fault prediction.

[0187] The aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy provided by the embodiments of this specification includes obtaining at least two monitoring parameters of the hydraulic system of the target device in a historical time interval, where the at least two monitoring parameters include a hydraulic oil quantity parameter; correcting the parameter value of the hydraulic oil quantity parameter to generate a characteristic value of the hydraulic oil quantity parameter; in the case of determining that the characteristic value belongs to a pre-generated parameter value interval corresponding to the hydraulic oil quantity parameter, updating the parameter value interval to generate a target parameter value interval; inputting the characteristic value into a pre-trained fault detection model for processing to generate a predicted value of the hydraulic oil quantity of the hydraulic system at the target moment; and determining a fault detection result of the hydraulic system at the target moment based on the predicted value of the hydraulic oil quantity and the target parameter value interval. By correcting the parameter value of the hydraulic oil quantity parameter, not only the hydraulic oil quantity in the hydraulic oil tank is considered, but also the hydraulic oil quantity in the hydraulic pipeline is considered, making the corrected hydraulic oil quantity closer to the actual oil quantity, with better authenticity and predictability. In addition, by updating the parameter value interval, it is beneficial to ensure the long-term effectiveness of the upper and lower limit thresholds of the interval, and can effectively reduce the false alarm rate. In addition, predicting the hydraulic oil quantity based on the fault detection model, so as to realize fault prediction of the hydraulic system, is beneficial to improving the accuracy and stability of the prediction result.

[0188] Figure 3 The flowchart of the processing procedure of an aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy provided by an embodiment of this specification is shown, which specifically includes the following steps.

[0189] Step 302: Obtain the historical data of the aircraft hydraulic system in a historical time interval.

[0190] Step 304: Extract the hydraulic oil quantity data segment before engine startup.

[0191] Step 306: Perform extrapolation correction and brake release correction on the parameter value of the hydraulic oil quantity parameter.

[0192] Step 308: Construct a parameter value interval corresponding to the hydraulic oil quantity parameter based on the three-sigma principle.

[0193] Step 310: For the parameter value of the new hydraulic oil quantity parameter, update the upper and lower limit values of the parameter value interval based on Bayesian estimation and in combination with the parameter value.

[0194] Step 312: Fault detection model training.

[0195] Step 314: Predict the hydraulic oil quantity through the fault detection model.

[0196] Figure 4 The flowchart showing the parameter value correction process of a hydraulic oil quantity parameter provided by an embodiment of this specification is specifically as follows.

[0197] Step 402: Obtain the historical data of the aircraft hydraulic system within the historical time interval.

[0198] Step 404: Determine the parameters of the extrapolation correction algorithm, including the reference shutdown duration and the empirical function of the hydraulic oil increase varying with the shutdown duration.

[0199] Step 406: Determine the parameters of the brake release correction algorithm, including the target hydraulic oil sudden increase amount when the brake is released.

[0200] Step 408: For the parameter value of the new hydraulic oil quantity parameter, perform extrapolation correction and brake release correction on the parameter value.

[0201] Step 410: Obtain the corrected hydraulic oil quantity.

[0202] Specifically, the corrected hydraulic oil quantity is the characteristic value of the corrected hydraulic oil quantity parameter.

[0203] Figure 5 The flowchart showing the update process of a parameter value interval provided by an embodiment of this specification is specifically as follows.

[0204] Step 502: Obtain the historical data of the aircraft hydraulic system within the historical time interval.

[0205] Specifically, the historical data contains multiple historical characteristic values of the hydraulic oil quantity parameter.

[0206] Step 504: Calculate the mean and standard deviation of the multiple historical characteristic values.

[0207] Step 506: Construct the parameter value interval corresponding to the hydraulic oil quantity parameter by using the three-sigma principle.

[0208] Step 508: For the parameter value of the new hydraulic oil quantity parameter, determine whether the parameter value is within the parameter value interval; if so, execute Step 510; if not, end.

[0209] Step 510: Use the upper and lower limit values of the parameter value range as prior information, use the eigenvalue and at least two historical eigenvalue among multiple historical eigenvalues as posterior information, and update the upper and lower limit values of the parameter value range by using Bayesian estimation.

[0210] Figure 6 FIG. 4 shows a flowchart of a fault detection process provided by an embodiment of the present specification, which specifically includes the following steps.

[0211] Step 602: Obtain historical data of the aircraft hydraulic system within a historical time range.

[0212] Step 604: Slide and divide the historical data into sample data and sample labels.

[0213] Step 606: Input the sample data into an encoder and a decoder for pre-training.

[0214] Step 608: Freeze the parameters of the encoder, input the sample data into the encoder and a prediction network for training, and generate a fault detection model.

[0215] Step 610: Combine at least two historical eigenvalues and a new eigenvalue into an input sample.

[0216] Step 612: Input the input sample into the fault detection model for prediction processing, and generate a predicted value of the hydraulic oil quantity at the target moment.

[0217] Step 614: Determine a fault detection result of the hydraulic system at the target moment based on the predicted value of the hydraulic oil quantity and the target parameter value range.

[0218] Corresponding to the above method embodiment, the present specification also provides an embodiment of an aircraft hydraulic oil leakage fault diagnosis and prediction device based on a two-step correction strategy. Figure 7 FIG. 5 shows a schematic structural diagram of an aircraft hydraulic oil leakage fault diagnosis and prediction device based on a two-step correction strategy provided by an embodiment of the present specification. As Figure 7 shown, the device includes:

[0219] An acquisition module 702, configured to acquire at least two monitoring parameters of the hydraulic system of the target device within a historical time range, where the at least two monitoring parameters include a hydraulic oil quantity parameter;

[0220] A correction module 704, configured to correct the parameter value of the hydraulic oil quantity parameter to generate an eigenvalue of the hydraulic oil quantity parameter;

[0221] An update module 706, configured to update the parameter value range to generate a target parameter value range when it is determined that the eigenvalue belongs to the parameter value range corresponding to the pre-generated hydraulic oil quantity parameter.

[0222] A processing module 708, configured to input the eigenvalue into a pre-trained fault detection model for processing to generate a predicted value of the hydraulic oil quantity of the hydraulic system at the target moment.

[0223] A determination module 710, configured to determine a fault detection result of the hydraulic system at the target moment based on the predicted value of the hydraulic oil quantity and the target parameter value range.

[0224] Optionally, the correction module 704 is further configured to:

[0225] Determine a first parameter value of the hydraulic oil quantity parameter of the hydraulic system at the current startup moment, and determine a target shutdown duration of the hydraulic system according to the current startup moment and the previous shutdown moment before the current startup moment.

[0226] Based on the target shutdown duration and the correlation between the shutdown duration and the hydraulic oil increase amount, correct the first parameter value to generate a second parameter value.

[0227] Optionally, the at least two detection parameters further include a historical shutdown duration and the hydraulic oil increase amount corresponding to the historical duration.

[0228] Correspondingly, the correction module 704 is further configured to:

[0229] Determine a reference shutdown duration of the hydraulic system according to multiple historical shutdown durations within the historical time interval.

[0230] Establish an empirical function between the shutdown duration and the hydraulic oil increase amount according to the corresponding relationship between the multiple historical shutdown durations and the multiple hydraulic oil increase amounts.

[0231] Based on the target shutdown duration, the reference shutdown duration, and the empirical function, correct the first parameter value.

[0232] Optionally, the at least two detection parameters further include a historical hydraulic oil sudden increase amount of the hydraulic system when the target device releases the parking brake.

[0233] Correspondingly, the correction module 704 is further configured to:

[0234] Determine whether the target device releases the parking brake before the engine of the target device starts.

[0235] If not, determine the target hydraulic oil sudden increase of the hydraulic system when the brake is released according to multiple historical hydraulic oil sudden increases within the historical time interval;

[0236] Correct the second parameter value based on the target hydraulic oil sudden increase.

[0237] Optionally, the aircraft hydraulic oil leakage fault diagnosis and prediction device based on the two-step correction strategy further includes a construction module configured to:

[0238] Determine multiple historical characteristic values of the hydraulic oil quantity parameter within the historical time interval;

[0239] Calculate the mean value and standard deviation corresponding to the multiple historical characteristic values;

[0240] Based on the mean value and the standard deviation, and using the three-sigma principle, construct a parameter value interval corresponding to the hydraulic oil quantity parameter.

[0241] Optionally, the update module 706 is further configured to:

[0242] Use the upper and lower limit values of the parameter value interval as prior information, use the characteristic value and at least two historical characteristic values among the multiple historical characteristic values as posterior information, and update the upper and lower limit values of the parameter value interval by using Bayesian estimation to generate a target parameter value interval.

[0243] Optionally, the fault detection model includes an encoder, a decoder, and a prediction network. The encoder and the decoder are respectively composed of two layers of long short-term memory networks, and the prediction network is composed of four layers of fully connected layers;

[0244] Correspondingly, the processing module 708 is further configured to:

[0245] Determine multiple historical characteristic values of the hydraulic oil quantity parameter within the historical time interval;

[0246] Use the characteristic value and at least two historical characteristic values among the multiple historical characteristic values as input data to input into the encoder, and perform encoding processing on the input data through the encoder to generate corresponding embedding vectors;

[0247] Input the embedding vectors into the prediction network for processing.

[0248] Optionally, the aircraft hydraulic oil leakage fault diagnosis and prediction device based on the two-step correction strategy further includes a training module configured to:

[0249] Perform sliding segmentation on the multiple historical characteristic values to generate sample data and sample labels;

[0250] Input the sample data into the encoder of the to-be-trained fault detection model for encoding to generate a first embedding vector;

[0251] Input the first embedding vector into the prediction network of the to-be-trained fault detection model for processing to generate a corresponding prediction result;

[0252] Calculate the first loss value of the to-be-trained fault detection model based on the sample label and the prediction result;

[0253] Calculate the first gradient value corresponding to the network parameters of the prediction network according to the chain rule of differentiation and in combination with the first loss value;

[0254] Update the network parameters according to the gradient descent method and in combination with the first gradient value;

[0255] When it is determined that the loss function of the to-be-trained fault detection model converges according to the first loss value, stop training to obtain the fault detection model.

[0256] Optionally, the training module is further configured to:

[0257] Input the sample data into the encoder of the initial fault detection model for encoding to generate a second embedding vector;

[0258] Input the second embedding vector into the decoder of the initial fault detection model for decoding to generate a reconstructed sample data corresponding to the sample data;

[0259] Calculate the second loss value of the initial fault detection model based on the sample data and the reconstructed sample data;

[0260] Calculate the second gradient values corresponding to the network parameters of the encoder and the decoder according to the chain rule of differentiation and in combination with the second loss value;

[0261] Update the network parameters of the encoder and the decoder according to the gradient descent method and in combination with the second gradient values;

[0262] When it is determined that the loss function of the initial fault detection model converges according to the second loss value, stop training to obtain the to-be-trained fault detection model.

[0263] The above is a schematic solution of an aircraft hydraulic oil leakage fault diagnosis and prediction device based on a two-step correction strategy according to this embodiment. It should be noted that the technical solution of the aircraft hydraulic oil leakage fault diagnosis and prediction device based on the two-step correction strategy belongs to the same concept as the above-mentioned technical solution of the aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy. For the details not described in the technical solution of the aircraft hydraulic oil leakage fault diagnosis and prediction device based on the two-step correction strategy, reference can be made to the description of the technical solution of the aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy above.

[0264] Figure 8 FIG. 4 shows a structural block diagram of a computing device 800 provided according to an embodiment of this specification. The components of the computing device 800 include, but are not limited to, a memory 810 and a processor 820. The processor 820 is connected to the memory 810 through a bus 830, and a database 850 is used to store data.

[0265] The computing device 800 further includes an access device 840, and the access device 840 enables the computing device 800 to communicate via one or more networks 860. Examples of these networks include the Public Switched Telephone Network (PSTN), Local Area Network (LAN), Wide Area Network (WAN), Personal Area Network (PAN), or a combination of communication networks such as the Internet. The access device 840 may include one or more of any type of wired or wireless network interfaces (e.g., Network Interface Card (NIC)), such as an IEEE802.11 Wireless Local Area Network (WLAN) wireless interface, a Worldwide Interoperability for Microwave Access (Wi-MAX) interface, an Ethernet interface, a Universal Serial Bus (USB) interface, a cellular network interface, a Bluetooth interface, a Near Field Communication (NFC) interface, and so on.

[0266] In an embodiment of this specification, the above components of the computing device 800 and Figure 8 other components not shown in FIG. 4 may also be connected to each other, for example, through a bus. It should be understood that Figure 8 the shown structural block diagram of the computing device is only for example purposes and is not a limitation on the scope of this specification. Those skilled in the art can add or replace other components as needed.

[0267] The computing device 800 can be any type of stationary or mobile computing device, including mobile computers or mobile computing devices (e.g., tablet computers, personal digital assistants, laptop computers, notebook computers, netbooks, etc.), mobile phones (e.g., smartphones), wearable computing devices (e.g., smartwatches, smart glasses, etc.) or other types of mobile devices, or stationary computing devices such as desktop computers or PCs. The computing device 800 can also be a mobile or stationary server.

[0268] Among them, the processor 820 is used to execute the following computer-executable instructions, and when the computer-executable instructions are executed by the processor, the steps of the above-mentioned aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy are implemented.

[0269] The above is a schematic solution of a computing device according to this embodiment. It should be noted that the technical solution of this computing device and the technical solution of the above-mentioned aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy belong to the same concept. For the detailed content not described in the technical solution of the computing device, reference can be made to the description of the technical solution of the above-mentioned aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy.

[0270] An embodiment of this specification also provides a computer-readable storage medium, which stores computer-executable instructions, and when the computer-executable instructions are executed by the processor, the steps of the above-mentioned aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy are implemented.

[0271] The above is a schematic solution of a computer-readable storage medium according to this embodiment. It should be noted that the technical solution of this storage medium and the technical solution of the above-mentioned aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy belong to the same concept. For the detailed content not described in the technical solution of the storage medium, reference can be made to the description of the technical solution of the above-mentioned aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy.

[0272] An embodiment of this specification also provides a computer program, wherein when the computer program is executed on a computer, the computer is made to execute the steps of the above-mentioned aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy.

[0273] The above is a schematic solution of a computer program according to this embodiment. It should be noted that the technical solution of this computer program and the technical solution of the above-mentioned aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy belong to the same concept. For the detailed content not described in the technical solution of the computer program, reference can be made to the description of the technical solution of the above-mentioned aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy.

[0274] The above describes specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve the desired result. Additionally, the processes depicted in the figures do not necessarily require the particular order or sequential order shown to achieve the desired result. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0275] The computer instructions include computer program code, which may be in the form of source code, object code, executable files, or some intermediate forms, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, mobile hard disks, magnetic disks, optical disks, computer memories, read-only memories (ROM), random access memories (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0276] It should be noted that, for the foregoing method embodiments, for the sake of simplicity of description, they are all expressed as a series of action combinations. However, those skilled in the art should know that the embodiments of this specification are not limited by the described action sequences, because according to the embodiments of this specification, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the embodiments of this specification.

[0277] In the above embodiments, the descriptions of the various embodiments have their own focuses. For the parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0278] The preferred embodiments of this specification disclosed above are only used to help explain this specification. The alternative embodiments do not describe all the details in detail, nor do they limit the invention to the specific embodiments described. Obviously, many modifications and variations can be made according to the content of the embodiments of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the embodiments of this specification, so that those skilled in the art can understand and utilize this specification well. This specification is only limited by the claims and their full scope and equivalents.

Claims

1. A method for diagnosing and predicting aircraft hydraulic oil leakage faults based on a two-step correction strategy, comprising: Acquire at least two monitoring parameters of the hydraulic system of the target device within a historical time interval, wherein the at least two monitoring parameters include a hydraulic oil volume parameter; According to the hydraulic oil increase amounts corresponding to the target downtime and the reference downtime, respectively, a parameter value of a hydraulic oil quantity parameter of the hydraulic system at the current start-up time is corrected once, and according to the target hydraulic oil sudden increase amount of the hydraulic system when the brake is released, a secondary correction is made to the parameter value to generate a characteristic value of the hydraulic oil quantity parameter; When it is determined that the characteristic value belongs to the pre-generated parameter value interval corresponding to the hydraulic oil quantity parameter, updating the parameter value interval based on the characteristic value to generate a target parameter value interval; Inputting the characteristic value into a pre-trained fault detection model for processing to generate a predicted value of the hydraulic oil volume of the hydraulic system at a target time; Based on the predicted value of the hydraulic oil quantity and the target parameter value interval, a fault detection result of the hydraulic system at the target time is determined.

2. The aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy according to claim 1 further comprises: Determining a target downtime duration of the hydraulic system according to the current start time and a downtime duration before the current start time; Based on the target downtime duration and the correlation between the downtime duration and the increased amount of hydraulic oil, the increased amount of hydraulic oil corresponding to the target downtime duration is determined.

3. According to the aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy of claim 2, the at least two monitoring parameters also include the historical downtime and the hydraulic oil increase amount corresponding to the historical downtime; Accordingly, the method further comprises: Determining a reference downtime duration of the hydraulic system according to a plurality of historical downtime durations within the historical time interval; Establishing an empirical function between the downtime and the amount of hydraulic oil increase according to the correspondence between the multiple historical downtimes and the multiple amounts of hydraulic oil increase; The hydraulic oil increase amounts corresponding to the target downtime duration and the reference downtime duration are determined based on the empirical function.

4. The aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy according to claim 2 or 3, wherein the at least two monitoring parameters further include the historical hydraulic oil surge increment of the hydraulic system when the target device releases the parking brake; Accordingly, the method further comprises: determining whether the target device releases a parking brake before the engine of the target device is started; If not, then the target hydraulic oil surge increment of the hydraulic system when the brake is released is determined according to a plurality of historical hydraulic oil surge increments within the historical time interval.

5. The aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy according to claim 1 further comprises: Determining a plurality of historical characteristic values ​​of the hydraulic oil volume parameter within the historical time interval; Calculate the mean and standard deviation corresponding to the multiple historical characteristic values; Based on the mean value and the standard deviation, a parameter value interval corresponding to the hydraulic oil quantity parameter is constructed using the three sigma principle.

6. The aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy according to claim 5, wherein the updating of the parameter value interval to generate the target parameter value interval comprises: The upper and lower limits of the parameter value interval are used as prior information, the characteristic value and at least two of the multiple historical characteristic values ​​are used as posterior information, and the upper and lower limits of the parameter value interval are updated using Bayesian estimation to generate a target parameter value interval.

7. According to the aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy of claim 1, the fault detection model comprises an encoder, a decoder and a prediction network, the encoder and the decoder are respectively composed of two layers of long short-term memory networks, and the prediction network is composed of four layers of fully connected layers; Accordingly, the inputting the characteristic value into a pre-trained fault detection model for processing includes: Determining a plurality of historical characteristic values ​​of the hydraulic oil volume parameter within the historical time interval; Inputting the feature value and at least two of the multiple historical feature values ​​as input data into the encoder, and encoding the input data by the encoder to generate a corresponding embedding vector; The embedding vector is input into the prediction network for processing.

8. The aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy according to claim 7 further comprises: Slidingly segmenting the multiple historical feature values ​​to generate sample data and sample labels; Inputting the sample data into the encoder of the fault detection model to be trained for encoding processing to generate a first embedding vector; Inputting the first embedding vector into the prediction network of the fault detection model to be trained for processing to generate a corresponding prediction result; Calculate a first loss value of the fault detection model to be trained based on the sample label and the prediction result; Calculate a first gradient value corresponding to a network parameter of the prediction network according to the chain derivative rule and in combination with the first loss value; Updating the network parameters according to a gradient descent method in combination with the first gradient value; When it is determined according to the first loss value that the loss function of the fault detection model to be trained converges, the training is stopped to obtain the fault detection model.

9. The aircraft hydraulic oil leakage fault diagnosis and prediction method based on the two-step correction strategy according to claim 8 further comprises: Inputting the sample data into the encoder of the initial fault detection model for encoding processing to generate a second embedding vector; Inputting the second embedding vector into the decoder of the initial fault detection model for decoding processing to generate reconstructed sample data corresponding to the sample data; Calculate a second loss value of the initial fault detection model based on the sample data and the reconstructed sample data; Calculate second gradient values ​​corresponding to network parameters of the encoder and the decoder according to the chain derivative rule and in combination with the second loss value; Update the network parameters of the encoder and the decoder according to the gradient descent method and in combination with the second gradient value; When it is determined according to the second loss value that the loss function of the initial fault detection model converges, the training is stopped to obtain the fault detection model to be trained.

10. An aircraft hydraulic oil leakage fault diagnosis and prediction device based on a two-step correction strategy, comprising: An acquisition module is configured to acquire at least two monitoring parameters of a hydraulic system of a target device within a historical time interval, wherein the at least two monitoring parameters include a hydraulic oil volume parameter; a correction module configured to perform a primary correction on a parameter value of a hydraulic oil quantity parameter of the hydraulic system at a current start-up time according to the hydraulic oil increase amounts corresponding to the target downtime duration and the reference downtime duration, respectively, and perform a secondary correction on the parameter value according to the target hydraulic oil sudden increase amount of the hydraulic system when the brake is released, so as to generate a characteristic value of the hydraulic oil quantity parameter; an updating module configured to, when it is determined that the characteristic value belongs to a parameter value interval corresponding to the pre-generated hydraulic oil quantity parameter, update the parameter value interval based on the characteristic value to generate a target parameter value interval; a processing module configured to input the characteristic value into a pre-trained fault detection model for processing, and generate a predicted value of the hydraulic oil volume of the hydraulic system at a target time; The determination module is configured to determine a fault detection result of the hydraulic system at the target time based on the predicted value of the hydraulic oil quantity and the target parameter value interval.

Images