CN116362142A - Aeroengine lubricating oil quantity prediction method, device, equipment and storage medium - Google Patents

Abstract

The invention relates to the field of aeroengines, in particular to a method for predicting the lubricating oil quantity of an aeroengine, which comprises the steps of constructing a regression mapping model of flight parameters and the lubricating oil quantity through a cyclic neural network algorithm, taking a lubricating oil quantity time sequence and a flying parameter time sequence as a prediction module of the lubricating oil quantity prediction model, carrying out wavelet decomposition, taking a corresponding scale function as training data, effectively reducing the sequence length of the training data, solving the problems of gradient elimination and gradient explosion in the long-time sequence training process, and having stronger fitting capability on the training data with time dependence, so that the prediction module can accurately reflect the correlation between the flying parameters and the lubricating oil quantity, and improving the training accuracy and efficiency, thereby monitoring a lubricating oil system more accurately and effectively.

Description

Aeroengine lubricating oil quantity prediction method, device, equipment and storage medium

technical field

The invention relates to the field of aero-engines, in particular to a method, device, equipment and storage medium for predicting the amount of lubricating oil in an aero-engine.

Background technique

Among aircraft engine failures, oil system failures account for a considerable proportion. Failure of the lubricating oil system can lead to engine shutdown, seriously affecting flight safety. At present, the research on the health monitoring of lubricating oil system mainly focuses on the monitoring of characteristic quantities such as lubricating oil temperature, lubricating oil consumption, abrasive particles in lubricating oil, and lubricating oil pressure. However, in practical applications, most of the fault monitoring of lubricating oil systems is still based on experience and qualitative analysis, which is not accurate enough and quantitative. Therefore, how to monitor the lubricating oil system and effectively predict the health status of the lubricating oil system is an urgent need research question.

The monitoring methods of the traditional lubricating oil system include low fuel volume alarm, post-flight manual monitoring, and dual-engine difference alarm. The change is affected by many factors, such as engine high-pressure rotor speed, aircraft attitude, flight speed and other factors. It is difficult to monitor the lubricating oil system accurately and effectively, which affects flight safety.

Contents of the invention

Based on this, the object of the present invention is to provide a method, device, equipment and storage medium for predicting the amount of lubricating oil in an aero-engine, and construct a regression mapping model of flight parameters and lubricating oil amount through a recurrent neural network algorithm as a predictive model for lubricating oil amount The prediction module, and the wavelet decomposition of the lubricating oil amount time series and the flight parameter time series, using the corresponding scale function as the training data, effectively reduces the sequence length of the training data, and can solve the problem of gradient disappearance and The gradient explosion problem has a stronger ability to fit time-dependent training data, so that the prediction module can accurately reflect the correlation between flight parameters and the amount of lubricating oil, which improves the training accuracy and efficiency, making it more accurate and effective Monitor the lubricating oil system.

In the first aspect, the embodiment of the present application provides a method for predicting the amount of lubricating oil in an aero-engine, comprising the following steps:

Obtaining record data corresponding to several flights, and constructing parameter time series sets corresponding to several flights, wherein the record data includes lubricating oil quantity record data and several flight parameter record data, and the parameter time series set includes lubricating oil quantity time series, and several flight parameter time series;

Obtaining a preset lubricating oil quantity prediction model, wherein the lubricating oil quantity prediction model includes a wavelet decomposition module and a prediction module to be trained;

Input the parameter time series sets corresponding to the several flights into the wavelet decomposition module for wavelet decomposition, and obtain the scaling function value sets corresponding to the several flights, wherein the scaling function value set includes the time series corresponding to each flight parameter The scale function value of , and the scale function value corresponding to the time series of lubricating oil quantity;

Input the scale function value sets corresponding to several flights into the prediction module to be trained for training to obtain a target lubricating oil quantity prediction model;

In response to the prediction instruction, obtain the flight parameter record data to be predicted, construct several flight parameter time series corresponding to the flight record data to be predicted, and input the several flight parameter time series corresponding to the flight record data to be predicted To the target lubricating oil quantity prediction model, obtain the lubricating oil quantity prediction result corresponding to the flight parameter record data to be predicted.

In a second aspect, the embodiment of the present application provides a device for predicting the amount of lubricating oil in an aero-engine, including:

The data acquisition module is used to obtain the record data corresponding to several flights, and construct the parameter time series set corresponding to several flights, wherein the record data includes record data of lubricating oil quantity and record data of several flight parameters, and the parameter time The sequence set includes the time series of lubricating oil quantity and several time series of flight parameters;

A model acquisition module, configured to obtain a preset lubricating oil quantity prediction model, wherein the lubricating oil quantity prediction model includes a wavelet decomposition module and a prediction module to be trained;

The wavelet decomposition module is configured to input the parameter time series sets corresponding to the several flights into the wavelet decomposition module for wavelet decomposition, and obtain the scaling function value sets corresponding to the several flights, wherein the scaling function value set includes The scaling function value corresponding to each flight parameter time series, and the scaling function value corresponding to the lubricating oil quantity time series;

A model training module, configured to input the scale function value sets corresponding to several flights into the prediction module to be trained for training to obtain a target lubricating oil quantity prediction model;

The lubricating oil quantity prediction module is used to respond to the prediction instruction, obtain the flight parameter record data to be predicted, construct several flight parameter time series corresponding to the flight record data to be predicted, and correspond the flight record data to be predicted Several time series of flight parameters are input to the target lubricating oil quantity prediction model, and the lubricating oil quantity prediction results corresponding to the recorded data of the flight parameters to be predicted are obtained.

In a third aspect, the embodiment of the present application provides a computer device, including: a processor, a memory, and a computer program stored on the memory and operable on the processor; the computer program is executed by the processor During execution, the steps of the method for predicting the amount of lubricating oil in an aeroengine as described in the first aspect are realized.

In a fourth aspect, an embodiment of the present application provides a storage medium, the storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method for predicting the amount of lubricating oil in an aeroengine as described in the first aspect are implemented .

In the embodiment of the present application, a method, device, device, and storage medium for predicting the amount of lubricating oil in an aeroengine are provided to construct a regression mapping model of flight parameters and lubricating oil amount through a cyclic neural network algorithm, as a prediction module of the lubricating oil amount prediction model , and decompose the lubricating oil amount time series and flight parameter time series by wavelet, and use the corresponding scale function as the training data, which effectively reduces the sequence length of the training data and can solve the problem of gradient disappearance and gradient explosion in the long-term sequence training process , the ability to fit time-dependent training data is stronger, so that the prediction module can accurately reflect the correlation between flight parameters and the amount of lubricating oil, improve the training accuracy and efficiency, and thus more accurately and effectively The oil system is monitored.

For better understanding and implementation, the present invention will be described in detail below in conjunction with the accompanying drawings.

Description of drawings

FIG. 1 is a schematic diagram of an application scenario of a method for predicting the amount of lubricating oil in an aeroengine provided by an embodiment of the present application;

Fig. 2 is a schematic flow chart of a method for predicting the amount of lubricating oil in an aero-engine provided by an embodiment of the present application;

FIG. 3 is a schematic flow diagram of S3 in the method for predicting the amount of lubricating oil in an aeroengine provided by an embodiment of the present application;

FIG. 4 is a schematic flow chart of S4 in the method for predicting the amount of lubricating oil in an aeroengine provided by an embodiment of the present application;

FIG. 5 is a schematic flow chart of S5 in the method for predicting the amount of lubricating oil in an aeroengine provided by an embodiment of the present application;

Fig. 6 is a schematic structural diagram of an aeroengine lubricating oil amount prediction device provided by an embodiment of the present application;

Fig. 7 is a schematic structural diagram of a computer device provided by an embodiment of the present application.

Detailed ways

Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numerals in different drawings refer to the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with aspects of the present application as recited in the appended claims.

The terminology used in this application is for the purpose of describing particular embodiments only, and is not intended to limit the application. As used in this application and the appended claims, the singular forms "a", "the", and "the" are intended to include the plural forms as well, unless the context clearly dictates otherwise. It should also be understood that the term "and/or" as used herein refers to and includes any and all possible combinations of one or more of the associated listed items.

It should be understood that although the terms first, second, third, etc. may be used in this application to describe various information, the information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of the present application, first information may also be called second information, and similarly, second information may also be called first information. Depending on the context, the words "if"/"if" as used herein may be interpreted as "at" or "when" or "in response to a determination".

The data sending end may be a computer device or a mobile terminal device, which is used to establish a network connection with the data receiving end, and is capable of encoding the data information sent to the data receiving end, and encoding the data information from the data receiving end. The data information sent by the data receiving end is analyzed.

The data receiving end may be a computer device or a mobile terminal device, which is used to establish a network connection with the data sending end, and is capable of encoding data information sent to the data sending end, and The data information sent by the data sending end is analyzed.

Please refer to Fig. 1, Fig. 1 is a schematic flow chart of a method for predicting the amount of lubricating oil in an aeroengine provided by an embodiment of the present application, the method includes the following steps:

S1: Obtain the recorded data corresponding to several flights, and construct the parameter time series set corresponding to several flights.

The subject of execution of the method for predicting the amount of lubricating oil in an aero-engine of the present application is the prediction device of the method for predicting the amount of lubricating oil in an aero-engine (hereinafter referred to as the prediction device). In an optional embodiment, the forecasting device may be a computer device, a server, or a server cluster formed by combining multiple computer devices.

The recorded data is to use QAR data recording equipment to read various data from the aircraft bus (such as ARINC 429 bus or AFDX bus), and according to the preset record configuration table, the data collected every second is frame-subframe- Store in the form of sub-slots, and use decoding software to convert the binary data stream into the engineering value recorded by the sensor as the recorded data.

The aviation engine lubricating oil system includes key components such as the lubricating oil tank, oil filter, oil pump, heat exchanger, etc. The lubricating oil in the lubricating oil tank is supplied to the parts that need to be cooled and lubricated through the engine-driven oil pump, and then the lubricating oil from the oil tank and gears is supplied by the oil return pump. The lubricating oil in the lubricating oil tank returns to the lubricating oil tank to form the circulation of the lubricating oil in the engine system. The lubricating oil quantity sensor is located in the lubricating oil tank, which can be divided into two types: capacitive oil quantity sensor and magnetic float oil level sensor. Capacitive sensors are excited by alternating current, and the sensor generates a direct current signal proportional to the liquid level. The magnetic float type sensor uses a magnetic reed switch to move the valve stem up and down according to the rise and fall of the oil level to obtain a record of the oil quantity as the record data of the lubricating oil quantity.

The flight parameter record data includes record data corresponding to several flight parameters, wherein the flight parameters include engine high pressure rotor speed parameters, low pressure rotor speed parameters, aircraft flight height parameters and flight attitude parameters, and the flight attitude parameters include roll Rotation angle, pitch angle, roll angle difference, pitch angle difference, roll angle second order difference and pitch angle second order difference.

In this embodiment, the forecasting device obtains record data corresponding to several flights, and constructs a parameter time series set corresponding to several flights, wherein the record data includes record data of lubricating oil quantity and record data of several flight parameters, the The parameter time series set includes the time series of lubricating oil quantity and several time series of flight parameters.

In an optional embodiment, the record data of lubricating oil quantity includes the original lubricating oil quantity corresponding to several moments and the benchmark lubricating oil quantity, the original lubricating oil quantity is the lubricating oil quantity recorded by the sensor at the beginning of time t , and the benchmark The amount of lubricating oil is the amount of lubricating oil recorded by the sensor at the end of time t, and the reference amount of lubricating oil is used to indicate the amount of lubricating oil when the aircraft is landing and returning oil. The lubricating oil amount time series includes the original lubricating oil amount and the reference lubricating oil amount corresponding to several moments, wherein the reference lubricating oil amount is the recorded lubricating oil amount at the end of several moments.

S2: Obtain a preset lubricating oil quantity prediction model, wherein the lubricating oil quantity prediction model includes a wavelet decomposition module and a prediction module to be trained.

In this embodiment, the prediction device obtains a preset lubricating oil quantity prediction model, wherein the lubricating oil quantity prediction model includes a wavelet decomposition module and a prediction module to be trained. The wavelet decomposition module includes several layers of wavelet decomposition layers. Due to the characteristics of multi-resolution analysis, wavelet decomposition can focus on any details of the signal for multi-resolution time-frequency domain analysis, which can improve the training accuracy of the prediction module.

The prediction module to be trained can use a long-term short-term memory network (LSTM). LSTM is a model constructed using a recurrent neural network algorithm, which can store historical information to form a long-term memory.

Please refer to Fig. 2, Fig. 2 is a schematic flow chart of a method for predicting the amount of lubricating oil in an aero-engine provided by another embodiment of the present application, which also includes steps S6~S7, and the steps S6~S7 are before step S3, as follows:

S6: Obtain lubricating oil temperature record data corresponding to several flights, and construct lubricating oil temperature time series corresponding to several flights.

When the engine is running, the lubricating oil is at different operating temperatures. Before the engine starts cold (usually the first start of the day), the lubricating oil temperature is basically the same as the outside air temperature. After the engine is started, the oil temperature of the lubricating oil is generally above 70°C. At this time, the value of the lubricating oil quantity will be affected by the thermal expansion effect.

In order to reduce the influence of the thermal expansion effect on the training of the lubricating oil prediction model, in this embodiment, the prediction equipment obtains the lubricating oil temperature record data corresponding to several flights, and constructs the lubricating oil temperature time series corresponding to several flights, which are used to analyze the The original lubricating oil amount corresponding to each moment in the lubricating oil amount time series is calibrated to improve the accuracy of the training of the prediction module of the lubricating oil amount prediction model.

S7: According to the lubricating oil temperature time series corresponding to several flights and the preset lubricating oil amount correction algorithm, respectively collect the parameter time series corresponding to several flights, and the lubricating oil amount corresponding to each moment in the lubricating oil amount time series The original lubricating oil amount is calibrated to obtain the calibrated lubricating oil amount corresponding to several times corresponding to several flights.

The algorithm for correcting the amount of lubricating oil is:

为第t个时刻对应的校准润滑油量，/>

为第t个时刻对应的滑油温度，/>

为第t个时刻对应的原始润滑油量，/>

为第t个时刻对应的基准润滑油量，

表示第t个时刻结束。In the formula,

is the calibration lubricating oil quantity corresponding to the tth moment, />

is the lubricating oil temperature corresponding to the tth moment, />

is the original lubricating oil amount corresponding to the tth moment, />

is the benchmark lubricating oil quantity corresponding to the tth moment,

Indicates the end of the tth moment.

In this embodiment, the prediction device collects the parameter time series corresponding to several flights according to the lubricating oil temperature time series corresponding to several flights and the preset lubricating oil amount correction algorithm, and the lubricating oil amount time The original lubricating oil quantity corresponding to each moment in the sequence is calibrated to obtain the calibrated lubricating oil quantity corresponding to several moments corresponding to several flights.

S3: Input the parameter time series sets corresponding to the several flights into the wavelet decomposition module for wavelet decomposition, and obtain the scaling function value sets corresponding to the several flights.

In this embodiment, the forecasting device inputs the parameter time series sets corresponding to the several flights into the wavelet decomposition module for wavelet decomposition, and obtains the scaling function value sets corresponding to the several flights, wherein the scaling function value The set includes the scaling function value corresponding to each flight parameter time series, and the scaling function value corresponding to the lubricating oil quantity time series.

The wavelet decomposition module includes several layers of wavelet decomposition layers. Please refer to FIG. 3. FIG. 3 is a schematic flow diagram of S3 in the method for predicting the amount of lubricating oil in an aeroengine provided by an embodiment of the present application, including step S31, as follows:

S31: Gather the parameter time series together, each flight parameter time series and lubricating oil amount time series as signals to be decomposed, and obtain scaling function value sets corresponding to several flights according to a preset wavelet decomposition algorithm.

The wavelet decomposition algorithm is:

为所述待分解信号，k为函数偏移量，j为缩放系数，表示当前小波分解层的层数，/>

为第j-1层的小波分解层的尺度函数系数，/>

为第j-1层的小波分解层输出的第t时刻对应的尺度函数值，/>

为第j-1层的小波分解层输出的第t时刻对应的小波函数系数，/>

为第j-1层的小波分解层输出的第t时刻对应的小波函数值，其中，尺度函数值以及小波函数值的表达式为：In the formula,

is the signal to be decomposed, k is the function offset, j is the scaling factor, indicating the number of layers of the current wavelet decomposition layer, />

is the scaling function coefficient of the wavelet decomposition layer of the j -1th layer, />

is the scaling function value corresponding to the tth moment output by the wavelet decomposition layer of the j -1th layer, />

is the wavelet function coefficient corresponding to the tth moment output by the wavelet decomposition layer of the j -1th layer, />

is the wavelet function value corresponding to the tth moment output by the wavelet decomposition layer of the j -1th layer, where the expressions of the scaling function value and the wavelet function value are:

为第j层的小波分解层输出的第t时刻对应的尺度函数值，/>

为第j+1层的小波分解层的低通滤波器系数，/>

为第j层的小波分解层输出的第t时刻对应的小波函数值，/>

为第j+1层的小波分解层的高通滤波器系数。In the formula,

is the scaling function value corresponding to the tth moment output by the wavelet decomposition layer of the jth layer, />

is the low-pass filter coefficient of the wavelet decomposition layer of the j +1th layer, />

is the wavelet function value corresponding to the t- th moment output by the wavelet decomposition layer of the j- th layer, />

is the high-pass filter coefficient of the wavelet decomposition layer of the j +1th layer.

In this embodiment, the forecasting device gathers the parameter time series, each flight parameter time series and lubricating oil amount time series as the signal to be decomposed, and obtains the scaling function value sets corresponding to several flights according to the preset wavelet decomposition algorithm , as the training data of the prediction module of the lubricating oil quantity prediction model.

The prediction equipment adopts the biorthogonal wavelet-based method, decomposes the signal to be decomposed by wavelet, and uses the scale function value after wavelet decomposition as training data, so that the input and output parameters of the prediction module used to train the lubricating oil quantity prediction model The sequence length is effectively reduced, which effectively improves the training accuracy and efficiency.

S4: Input the scaling function value sets corresponding to several flights into the prediction module to be trained for training to obtain a target lubricating oil quantity prediction model.

In this embodiment, the prediction device inputs the scaling function value sets corresponding to several flights into the prediction module to be trained for training to obtain a target lubricating oil quantity prediction model.

Please refer to FIG. 4. FIG. 4 is a schematic flow diagram of S4 in the method for predicting the amount of lubricating oil in an aeroengine provided by an embodiment of the present application, including steps S41 to S42, as follows:

S41: Using the scale function values corresponding to several flight parameter time series in the scale function value set as input data, according to the cyclic neural network algorithm, obtain the output gates corresponding to several times of the flight, and use the same flight The output gates corresponding to several corresponding times are combined to construct the scale function value corresponding to the predicted lubricating oil amount time series corresponding to several flights.

In this embodiment, the forecasting device normalizes the scale function values corresponding to several flight parameter time series in the set of scale function values as input data, and obtains several flight correspondences according to the cyclic neural network algorithm. The output gates corresponding to several moments, wherein, the process of the recurrent neural network algorithm is as follows:

为第t个时刻对应的遗忘门，/>

为遗忘门的权重矩阵参数，/>

为第t-1个时刻对应的输出门，/>

为第t个时刻对应的输入数据，/>

为遗忘门的偏置参数，σ()为激活函数，/>

为第t个时刻对应的输入门，/>

为输入门的权重矩阵参数，/>

为输入门的权重矩阵，/>

为第t个时刻对应的单元状态，/>

为第t个时刻对应的单元状态的变化量，/>

为单元状态的权重矩阵参数，/>

为单元状态的偏置参数，tanh为双曲正切函数。In the formula,

is the forgetting gate corresponding to the tth moment, />

is the weight matrix parameter of the forget gate, />

is the output gate corresponding to the t -1th moment, />

is the input data corresponding to the tth moment, />

is the bias parameter of the forget gate, σ() is the activation function, />

is the input gate corresponding to the tth moment, />

is the weight matrix parameter of the input gate, />

is the weight matrix of the input gate, />

is the unit state corresponding to the tth moment, />

is the change amount of the unit state corresponding to the tth moment, />

is the weight matrix parameter of the cell state, />

is the bias parameter of the unit state, and tanh is the hyperbolic tangent function.

The output gates are:

为第t-1个时刻对应的输出门，/>

为输出门的权重矩阵参数，/>

为输出门的偏置参数。In the formula,

is the output gate corresponding to the t -1th moment, />

is the weight matrix parameter of the output gate, />

is the bias parameter of the output gate.

The prediction device combines the output gates corresponding to several times corresponding to the same flight to construct the scale function value corresponding to the time series of predicted lubricating oil amount corresponding to several flights.

S42: Calculate the mean square error data according to the scaling function value corresponding to the predicted lubricating oil amount time series corresponding to the same flight and the scaling function value corresponding to the lubricating oil amount time series, and calculate the mean square error data according to the mean square error data. The prediction module to be trained is trained to obtain a target prediction module, and the target prediction module is combined with the wavelet decomposition module to obtain a target lubricating oil quantity prediction model.

In this embodiment, the prediction device calculates the mean square error data according to the scale function value corresponding to the predicted lubricating oil quantity time series corresponding to the same flight and the scaling function value corresponding to the lubricating oil quantity time series, and according to the mean square For error data, the prediction module to be trained is trained to obtain a target prediction module, and the target prediction module is combined with a wavelet decomposition module to obtain a target lubricating oil quantity prediction model.

The regressive mapping model of flight parameters and lubricating oil volume is constructed by the cyclic neural network algorithm. As the prediction module of the lubricating oil volume prediction model, it can solve the problem of gradient disappearance and gradient explosion in the long-term sequence training process, and the training data with time dependence The fitting ability is stronger, so that the prediction module can accurately reflect the correlation between the flight parameters and the amount of lubricating oil, so as to monitor the lubricating oil system more accurately and effectively.

S5: In response to the prediction instruction, obtain the flight parameter record data to be predicted, construct a number of flight parameter time series corresponding to the flight record data to be predicted, and set the flight parameter time series corresponding to the flight record data to be predicted The sequence is input to the target lubricating oil quantity prediction model to obtain the lubricating oil quantity prediction result corresponding to the flight parameter record data to be predicted.

The prediction instruction is sent by the user and received by the prediction device.

The prediction device responds to the prediction instruction, obtains the flight parameter record data to be predicted, constructs a number of flight parameter time series corresponding to the flight record data to be predicted, and sets the flight parameter time series corresponding to the flight record data to be predicted The sequence is input into the target lubricating oil quantity prediction model to obtain the lubricating oil quantity prediction result corresponding to the flight parameter record data to be predicted, and display the lubricating oil quantity prediction result on a preset display interface.

Please refer to Fig. 5. Fig. 5 is a schematic flow diagram of S5 in the method for predicting the amount of lubricating oil in an aeroengine provided by an embodiment of the present application, including steps S51 to S52, as follows:

S51: Input the time series of several flight parameters corresponding to the flight record data to be predicted into the wavelet decomposition module in the target lubricating oil quantity prediction model, and obtain the time of each flight parameter corresponding to the flight record data to be predicted The scale function value corresponding to the sequence.

In this embodiment, the prediction device inputs the time series of several flight parameters corresponding to the flight record data to be predicted into the wavelet decomposition module in the target lubricating oil quantity prediction model to obtain the flight record data to be predicted For the scaling function values corresponding to the time series of each flight parameter, the specific method can refer to step S31, which will not be repeated here.

S52: Input the scaling function value corresponding to each flight parameter time series corresponding to the flight record data to be predicted to the prediction module in the target lubricating oil quantity prediction model, and obtain the prediction corresponding to the flight record data to be predicted The scaling function value corresponding to the lubricating oil amount time series is subjected to differential processing on the scaling function value corresponding to the predicted lubricating oil amount time series corresponding to the flight record data to be predicted, and used as the lubricating oil amount prediction result.

In this embodiment, the prediction device inputs the scaling function values corresponding to the flight parameter time series corresponding to the flight record data to be predicted to the prediction module in the target lubricating oil quantity prediction model, and obtains the to-be-predicted The scale function value corresponding to the predicted lubricating oil quantity time series corresponding to the flight record data.

The forecasting device performs difference processing on the scale function value corresponding to the predicted lubricating oil amount time series corresponding to the flight record data to be predicted, and restores the scale function value to the time series corresponding to the flight record data to be predicted The lubricating oil quantity forecast data is used as the lubricating oil quantity forecasting result, so as to monitor the lubricating oil system accurately and effectively.

Please refer to Fig. 6. Fig. 6 is a schematic structural diagram of an aero-engine lubricating oil quantity prediction device provided by an embodiment of the present application, which can realize all or part of the aero-engine lubricating oil quantity prediction device through software, hardware or a combination of the two , the device 6 includes:

The data acquisition module 61 is used to obtain the record data corresponding to several flights, and construct the parameter time series set corresponding to several flights, wherein the record data includes record data of lubricating oil quantity and record data of several flight parameters, and the parameters The time series set includes the time series of lubricating oil quantity and several time series of flight parameters;

A model acquisition module 62, configured to obtain a preset lubricating oil quantity prediction model, wherein the lubricating oil quantity prediction model includes a wavelet decomposition module and a prediction module to be trained;

The wavelet decomposition module 63 is configured to input the parameter time series sets corresponding to the several flights into the wavelet decomposition module for wavelet decomposition, and obtain the scaling function value sets corresponding to the several flights, wherein the scaling function value set Including the scaling function value corresponding to each flight parameter time series, and the scaling function value corresponding to the lubricating oil quantity time series;

A model training module 64, configured to input the scale function value sets corresponding to several flights into the prediction module to be trained for training to obtain a target lubricating oil quantity prediction model;

The lubricating oil quantity prediction module 65 is used to respond to the prediction instruction, obtain the flight parameter record data to be predicted, construct several flight parameter time series corresponding to the flight record data to be predicted, and store the flight record data to be predicted The corresponding time series of several flight parameters are input to the target lubricating oil quantity prediction model, and the lubricating oil quantity prediction results corresponding to the recorded data of the flight parameters to be predicted are obtained.

In this embodiment, through the data acquisition module, the record data corresponding to several flights are obtained, and the parameter time series sets corresponding to several flights are constructed, wherein the record data includes lubricating oil quantity record data and several flight parameter record data , the parameter time series set includes a lubricating oil quantity time series, and several flight parameter time series; through the model acquisition module, a preset lubricating oil quantity prediction model is obtained, wherein the lubricating oil quantity prediction model includes a wavelet decomposition module And the prediction module to be trained; through the wavelet decomposition module, the parameter time series sets corresponding to the several flights are input into the wavelet decomposition module to perform wavelet decomposition, and the scaling function value sets corresponding to several flights are obtained, wherein the The scale function value set includes the scale function value corresponding to each flight parameter time series, and the scale function value corresponding to the lubricating oil amount time series; through the model training module, the scale function value set corresponding to several flights is input into the Carry out training in the prediction module to be trained to obtain the target lubricating oil quantity prediction model; through the lubricating oil quantity prediction module, in response to the prediction instruction, obtain the flight parameter record data to be predicted, and construct a number of data corresponding to the flight record data to be predicted time series of flight parameters, input the time series of flight parameters corresponding to the flight record data to be predicted into the target lubricating oil quantity prediction model, and obtain the lubricating oil quantity prediction corresponding to the flight parameter record data to be predicted result. Construct the regression mapping model of flight parameters and lubricating oil quantity through the cyclic neural network algorithm, as the prediction module of the lubricating oil quantity prediction model, and perform wavelet decomposition on the time series of lubricating oil quantity and flight parameter time series, and use the corresponding scaling function as training data, which effectively reduces the sequence length of training data, can solve the problem of gradient disappearance and gradient explosion in the long-term sequence training process, and has a stronger ability to fit time-dependent training data, so that the prediction module can accurately reflect the flight parameters. The correlation with the amount of lubricating oil improves the training accuracy and efficiency, so that the lubricating oil system can be monitored more accurately and effectively.

Please refer to FIG. 7. FIG. 7 is a schematic structural diagram of a computer device provided by an embodiment of the present application. The computer device 7 includes: a processor 71, a memory 72, and a computer program 73 stored on the memory 72 and operable on the processor 71 The computer device can store a plurality of instructions, and the instructions are suitable for being loaded by the processor 71 and executing the method steps of the above-mentioned FIGS. 1 to 5. The specific execution process can refer to the specific description of FIGS.

Wherein, the processor 71 may include one or more processing cores. Processor 71 uses various interfaces and lines to connect various parts in the server, by running or executing instructions, programs, code sets or instruction sets stored in memory 72, and calling data in memory 72, to execute aeroengine lubricating oil quantity Various functions and data processing of the predicting device 6, optionally, the processor 71 can use digital signal processing (Digital Signal Processing, DSP), field-programmable gate array (Field-Programmable Gate Array, FPGA), programmable logic array (ProgrambleLogic Array, PLA) in at least one form of hardware. The processor 71 may integrate one or a combination of a central processing unit 71 (Central Processing Unit, CPU), an image processor 71 (Graphics Processing Unit, GPU), a modem, and the like. Among them, the CPU mainly processes the operating system, user interface and application programs, etc.; the GPU is used for rendering and drawing the content that needs to be displayed on the touch screen; the modem is used for processing wireless communication. It can be understood that, the above-mentioned modem may not be integrated into the processor 71, but may be realized by a single chip.

Wherein, the memory 72 may include a random access memory 72 (Random Access Memory, RAM), and may also include a read-only memory 72 (Read-Only Memory). Optionally, the memory 72 includes a non-transitory computer-readable storage medium (non-transitory computer-readable storage medium). Memory 72 may be used to store instructions, programs, codes, sets of codes, or sets of instructions. The memory 72 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing the operating system, instructions for at least one function (such as touch instructions, etc.), and instructions for implementing the above-mentioned various method embodiments. Instructions, etc.; the storage data area can store data, etc. involved in the above method embodiments. Optionally, the memory 72 may also be at least one storage device located away from the aforementioned processor 71 .

The embodiment of the present application also provides a storage medium, the storage medium can store a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the above-mentioned method steps in Figure 1 to Figure 5, the specific execution process can be referred to Figure 1 to FIG. 5 are not repeated here.

Those skilled in the art can clearly understand that for the convenience and brevity of description, only the division of the above-mentioned functional units and modules is used for illustration. In practical applications, the above-mentioned functions can be assigned to different functional units, Completion of modules means that the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiment may be integrated into one processing unit, or each unit may exist separately physically, or two or more units may be integrated into one unit, and the above-mentioned integrated units may adopt hardware It can also be implemented in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the protection scope of the present application. For the specific working processes of the units and modules in the above system, reference may be made to the corresponding processes in the aforementioned method embodiments, and details will not be repeated here.

In the above-mentioned embodiments, the descriptions of each embodiment have their own emphases, and for parts that are not detailed or recorded in a certain embodiment, refer to the relevant descriptions of other embodiments.

Those skilled in the art can appreciate that the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed by hardware or software depends on the specific application of the technical solution and the design constraint algorithm. Skilled artisans may use different methods to implement the described functions for each specific application, but such implementation should not be regarded as exceeding the scope of the present invention.

In the embodiments provided in the present invention, it should be understood that the disclosed apparatus/terminal equipment and method may be implemented in other ways. For example, the device/terminal device embodiments described above are only illustrative. For example, the division of the modules or units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units Or components may be combined or may be integrated into another system, or some features may be omitted, or not implemented. In another point, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, and the indirect coupling or communication connection of devices or units may be in electrical, mechanical or other forms.

The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, each unit may exist separately physically, or two or more units may be integrated into one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

If the integrated module/unit is realized in the form of a software function unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention realizes all or part of the processes in the methods of the above embodiments, and can also be completed by instructing related hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer When the program is executed by the processor, the steps in the above-mentioned various method embodiments can be realized. Wherein, the computer program includes computer program code, and the computer program code may be in the form of source code, object code, executable file or some intermediate form.

The present invention is not limited to the above-mentioned embodiments, if the various changes or deformations of the present invention do not depart from the spirit and scope of the present invention, if these changes and deformations belong to the claims of the present invention and the equivalent technical scope, then the present invention is also It is intended that such modifications and variations are included.

Claims (10)

1. The method for predicting the lubricating oil quantity of the aero-engine is characterized by comprising the following steps of:

obtaining recording data corresponding to a plurality of flights, and constructing a parameter time sequence set corresponding to the plurality of flights, wherein the recording data comprises lubricating oil quantity recording data and a plurality of flight parameter recording data, and the parameter time sequence set comprises a lubricating oil quantity time sequence and a plurality of flight parameter time sequences;

obtaining a preset lubricating oil mass prediction model, wherein the lubricating oil mass prediction model comprises a wavelet decomposition module and a prediction module to be trained;

inputting the parameter time series set corresponding to the plurality of times of flight into the wavelet decomposition module to perform wavelet decomposition to obtain a scale function value set corresponding to the plurality of times of flight, wherein the scale function value set comprises scale function values corresponding to each flight parameter time series and scale function values corresponding to the lubricating oil quantity time series;

Inputting the scale function value sets corresponding to the plurality of flights into the prediction module to be trained for training to obtain a target lubricating oil quantity prediction model;

and responding to a prediction instruction, obtaining flight parameter record data to be predicted, constructing a plurality of flight parameter time sequences corresponding to the flight record data to be predicted, inputting the plurality of flight parameter time sequences corresponding to the flight record data to be predicted into the target lubricating oil quantity prediction model, and obtaining a lubricating oil quantity prediction result corresponding to the flight parameter record data to be predicted.

2. The method for predicting the amount of lubricating oil for an aircraft engine according to claim 1, wherein: the oil quantity recording data comprise original oil quantity and reference oil quantity corresponding to a plurality of moments, the oil quantity time sequence comprises the original oil quantity and the reference oil quantity corresponding to the plurality of moments, and the reference oil quantity is the recorded oil quantity when the plurality of moments are finished.

3. The method for predicting the lubricating oil quantity of an aeroengine according to claim 2, wherein before inputting the parameter time series set corresponding to the plurality of flights into the wavelet decomposition module to perform wavelet decomposition to obtain the scale function value set corresponding to the plurality of flights, the method further comprises the steps of:

Obtaining lubricating oil temperature record data corresponding to a plurality of flights, and constructing a lubricating oil temperature time sequence corresponding to the plurality of flights, wherein the lubricating oil temperature time sequence comprises lubricating oil temperatures corresponding to a plurality of moments;

according to the time series of the temperature of the lubricating oil corresponding to the plurality of flights and a preset lubricating oil quantity correction algorithm, the parameter time series corresponding to the plurality of flights are concentrated, the original lubricating oil quantity corresponding to each moment in the lubricating oil quantity time series is calibrated, and the calibrated lubricating oil quantity corresponding to the plurality of moments corresponding to the plurality of flights is obtained, wherein the lubricating oil quantity correction algorithm is as follows:

in the method, in the process of the invention,

is the firsttCalibrated amounts of lubricating oil for the respective moments +.>

Is the firsttThe temperature of the lubricating oil corresponding to each moment,

is the firsttOriginal lubricating oil quantity corresponding to each moment +.>

Is the firsttReference lubricating oil quantity corresponding to each moment +.>

Represent the firsttThe end of each moment.

4. The method for predicting the amount of lubricating oil for an aircraft engine according to claim 3, wherein: the wavelet decomposition module comprises a plurality of wavelet decomposition layers;

inputting the parameter time sequence set corresponding to the plurality of flights into the wavelet decomposition module to perform wavelet decomposition to obtain a scale function value set corresponding to the plurality of flights, wherein the method comprises the following steps of:

Taking the parameter time sequence set, each flight parameter time sequence and the lubricating oil quantity time sequence as signals to be decomposed, and obtaining a scale function value set corresponding to a plurality of flights according to a preset wavelet decomposition algorithm, wherein the wavelet decomposition algorithm is as follows:

in the method, in the process of the invention,

for the signal to be decomposed,kas a function of the amount of the offset,jfor scaling factor, the number of layers of the current wavelet decomposition layer is indicated, < >>

Is the firstj-scale function coefficients of wavelet decomposition layer of layer-1, < ->

Is the firstj-layer 1 wavelet decomposition layer outputtScale function value corresponding to moment, < >>

Is the firstj-layer 1 wavelet decomposition layer outputtWavelet function coefficient corresponding to time, +.>

Is the firstj-layer 1 wavelet decomposition layer outputtThe wavelet function value corresponding to the moment, wherein the scale function value and the wavelet function value are expressed as follows:

in the method, in the process of the invention,

is the firstjWavelet decomposition layer output of layertScale function value corresponding to moment, < >>

Is the firstjLow-pass filter coefficients of wavelet decomposition layer of +1 layer, +.>

Is the firstjWavelet decomposition layer output of layertWavelet function value corresponding to time, < >>

Is the firstjHigh pass filter coefficients of the wavelet decomposition layer of +1 layers.

5. The method for predicting the amount of lubricating oil for an aircraft engine according to claim 4, wherein: the prediction module to be trained is a model constructed by adopting a cyclic neural network algorithm, wherein the cyclic neural network algorithm process is as follows:

In the method, in the process of the invention,

is the firsttForgetting door corresponding to each moment +.>

Weight matrix parameters for forgetting gate, +.>

Is the firstt-1 output gate corresponding to moment, < ->

Is the firsttInput data corresponding to the respective time instant +.>

Sigma () is the activation function for the bias parameters of the forgetting gate, +.>

Is the firsttInput gate corresponding to each moment +.>

Weight matrix parameters for input gates, < +.>

As a weight matrix of the input gates,

is the firsttThe cell states corresponding to the respective moments +.>

Is the firsttThe amount of change of the state of the cell corresponding to the moment, < + >>

Weight matrix parameters for cell states, +.>

Is a bias parameter for the cell state, and tanh is a hyperbolic tangent function.

6. The method for predicting the amount of lubricating oil of an aeroengine according to claim 5, wherein the step of inputting the scale function value sets corresponding to the plurality of flights into the prediction module to be trained to perform training to obtain a target prediction model of the amount of lubricating oil comprises the steps of:

the scale function values corresponding to the time series of the flight parameters are collected to be used as input data, output doors corresponding to the time corresponding to the flight times are obtained according to the cyclic neural network algorithm, the output doors corresponding to the time corresponding to the same flight time are combined, and the scale function values corresponding to the time series of the predicted lubricating oil quantity corresponding to the flight times are constructed, wherein the output doors are:

In the method, in the process of the invention,

is the firstt-1 output gate corresponding to moment, < ->

For the weight matrix parameters of the output gate, < +.>

The bias parameters of the output gate;

and calculating mean square error data according to the scale function value corresponding to the predicted lubricating oil mass time sequence and the scale function value corresponding to the lubricating oil mass time sequence corresponding to the same flight, training the prediction module to be trained according to the mean square error data to obtain a target prediction module, and combining the target prediction module with a wavelet decomposition module to obtain a target lubricating oil mass prediction model.

7. The method for predicting the amount of lubricating oil of an aeroengine according to claim 6, wherein the step of inputting the plurality of time series of flight parameters corresponding to the flight record data to be predicted to the target lubricating oil amount prediction model to obtain the predicted result of the amount of lubricating oil corresponding to the flight record data to be predicted includes the steps of:

inputting a plurality of flight parameter time sequences corresponding to the flight record data to be predicted into a wavelet decomposition module in the target lubricating oil quantity prediction model to obtain scale function values corresponding to the flight parameter time sequences corresponding to the flight record data to be predicted;

And inputting the scale function values corresponding to the time series of each flight parameter corresponding to the flight record data to be predicted into a prediction module in the target lubricating oil quantity prediction model, obtaining the scale function values corresponding to the time series of the predicted lubricating oil quantity corresponding to the flight record data to be predicted, and performing difference processing on the scale function values corresponding to the time series of the predicted lubricating oil quantity corresponding to the flight record data to be predicted to serve as the lubricating oil quantity prediction result.

8. An aeroengine lubrication oil quantity prediction device, comprising:

the system comprises a data acquisition module, a data processing module and a data processing module, wherein the data acquisition module is used for acquiring record data corresponding to a plurality of flights and constructing a parameter time sequence set corresponding to the plurality of flights, the record data comprises lubricating oil quantity record data and a plurality of flight parameter record data, and the parameter time sequence set comprises a lubricating oil quantity time sequence and a plurality of flight parameter time sequences;

the model acquisition module is used for acquiring a preset lubricating oil quantity prediction model, wherein the lubricating oil quantity prediction model comprises a wavelet decomposition module and a prediction module to be trained;

the wavelet decomposition module is used for inputting the parameter time series set corresponding to the plurality of times of flight into the wavelet decomposition module to carry out wavelet decomposition, so as to obtain the scale function value set corresponding to the plurality of times of flight, wherein the scale function value set comprises the scale function value corresponding to each flight parameter time series and the scale function value corresponding to the lubricating oil quantity time series;

The model training module is used for inputting the scale function value sets corresponding to the plurality of flights into the prediction module to be trained for training to obtain a target lubricating oil quantity prediction model;

the lubrication oil quantity prediction module is used for responding to a prediction instruction, obtaining flight parameter record data to be predicted, constructing a plurality of flight parameter time sequences corresponding to the flight record data to be predicted, inputting the plurality of flight parameter time sequences corresponding to the flight record data to be predicted into the target lubrication oil quantity prediction model, and obtaining a lubrication oil quantity prediction result corresponding to the flight parameter record data to be predicted.

9. A computer device comprising a processor, a memory and a computer program stored in the memory and executable on the processor, the processor implementing the steps of the method for predicting the amount of lubricating oil of an aeroengine as claimed in any one of claims 1 to 7 when the computer program is executed by the processor.

10. A storage medium storing a computer program which, when executed by a processor, implements the steps of the aeroengine lubrication oil quantity prediction method of any of claims 1 to 7.

Images