Fuel injector failure early warning method, device, equipment and storage medium

Abstract

The application discloses a fuel injector fault early warning method, device, equipment and storage medium, and relates to the technical field of ship fault diagnosis. The fuel injector fault early warning method comprises the following steps: acquiring current operation state data and historical operation state data of a fuel injector, and constructing a fuel injector state estimation model based on a multivariate state estimation technology according to the historical operation state data; judging whether the current operation condition of the fuel injector is a steady state condition according to the current operation state data, and if yes, determining the similarity of the current operation state data of the fuel injector according to the current operation state data and the fuel injector state estimation model based on the multivariate state estimation technology; acquiring a current dynamic threshold value, judging whether the operation state of the fuel injector is faulty according to the current dynamic threshold value and the similarity of the current operation state data of the fuel injector, and judging whether the current dynamic threshold value is updated. The method can realize early warning before the operation state of the fuel injector is faulty, and improve the equipment maintenance efficiency and reliability.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of ship fault diagnosis, and particularly relates to an oil injector fault early warning method, device, equipment and storage medium. BACKGROUND

[0002] The high-pressure common rail technology of a marine diesel engine is the third technical leap in the development of diesel engines. Compared with the traditional mechanical fuel system, the high-pressure common rail system has the characteristics of high injection pressure and good environmental protection, and has the advantages of flexible adjustment of key indicators such as injection timing, cycle oil volume and injection duration. Among them, the electronically controlled injector, as the core component of the high-pressure common rail system, plays a major role in controlling the injection time and adjusting the injection duration. The performance of the electronically controlled injector will directly affect the in-cylinder combustion process of the diesel engine. If the electronically controlled injector fails, it will affect the cycle oil volume and atomization performance, making the in-cylinder combustion unable to meet the normal working requirements of the diesel engine, and in severe cases, it will affect the operation of the diesel engine and cause shutdown.

[0003] However, the existing diesel engine design does not have an online early warning function for the electronically controlled injector, so it cannot provide early warning before the electronically controlled injector fails, resulting in low equipment maintenance efficiency, high maintenance cost, poor equipment reliability and other problems. SUMMARY

[0004] The present application provides an oil injector fault early warning method, which aims to overcome the technical problems of the prior art that the current oil injector cannot provide early warning before failure, resulting in low equipment maintenance efficiency, high maintenance cost and poor equipment reliability. Another object of the present application is to provide an oil injector fault early warning device. A third object of the present application is to provide an electronic device. A fourth object of the present application is to provide a computer-readable storage medium.

[0005] Technical solution: The oil injector fault early warning method provided by the present application comprises:

[0006] Obtaining current running state data and historical running state data of the oil injector, and constructing an oil injector state estimation model based on a multivariate state estimation technology according to the historical running state data;

[0007] According to the current running state data, it is determined whether the current running condition of the oil injector is a steady state condition. If so, the similarity of the current running state data of the oil injector is determined according to the current running state data and the oil injector state estimation model based on the multivariate state estimation technology;

[0008] Obtaining a current dynamic threshold, judging whether the operation state of the fuel injector is faulty according to the similarity between the current dynamic threshold and current operation state data of the fuel injector, and judging whether to update the current dynamic threshold.

[0009] In some embodiments, the constructing the fuel injector state estimation model based on the multivariate state estimation technology according to the historical operation state data comprises:

[0010] Screening historical steady state working condition data from the historical operation state data;

[0011] Establishing the fuel injector state estimation model based on the multivariate state estimation technology according to the historical steady state working condition data.

[0012] In some embodiments, the establishing the fuel injector state estimation model based on the multivariate state estimation technology according to the historical steady state working condition data comprises:

[0013] Preprocessing the historical steady state working condition data to obtain a test matrix and a training matrix;

[0014] Generating a model observation matrix based on the training matrix;

[0015] Establishing the fuel injector state estimation model based on the multivariate state estimation technology according to the test matrix and the model observation matrix.

[0016] In some embodiments, the generating the model observation matrix based on the training matrix comprises:

[0017] Determining a health matrix according to the training matrix;

[0018] Calculating a model nonlinear operator based on the Euclidean distance;

[0019] Determining a model weight coefficient according to the nonlinear operator and the health matrix;

[0020] Generating the model observation matrix according to the model weight coefficient.

[0021] In some embodiments, the screening the historical steady state working condition data from the historical operation state data comprises:

[0022] Judging whether the historical operation state data is steady state working condition according to the historical operation state data, eliminating non-steady state working condition data in the historical operation state data, and retaining steady state working condition data.

[0023] In some embodiments, the judging the steady state working condition comprises:

[0024] acquire operation state data of the fuel injector; wherein the operation state data comprises fuel temperature data of the fuel injector and rail pressure data of the fuel injector;

[0025] determine whether the rail pressure of the fuel injector changes according to the rail pressure data of the fuel injector, and if so, further determine whether monotonicity of the fuel temperature of the fuel injector changes, and if so, determine that the current working condition is a steady state working condition.

[0026] In some embodiments, the determining whether the operation state of the fuel injector fails according to the similarity of the current dynamic threshold value and the current operation state data of the fuel injector, and determining whether to update the current dynamic threshold value, comprises:

[0027] comparing the similarity of the current operation state data of the fuel injector with the current dynamic threshold value;

[0028] if the similarity of the current operation state data of the fuel injector is greater than the current dynamic threshold value, determining that the current operation state data of the fuel injector is normal, and determining whether to update the current dynamic threshold value;

[0029] if the similarity of the current operation state data of the fuel injector is less than or equal to the current dynamic threshold value, determining that the current operation state data of the fuel injector is abnormal, and acquiring a time sequence of the current operation state data of the fuel injector; determining whether the operation state of the fuel injector fails according to the time sequence of the current operation state data of the fuel injector.

[0030] In some embodiments, the determining whether to update the current dynamic threshold value, comprises:

[0031] comparing the similarity of the current operation state data of the fuel injector with a static threshold value;

[0032] if the similarity of the current operation state data of the fuel injector is greater than the static threshold value, maintaining the current dynamic threshold value;

[0033] if the similarity of the current operation state data of the fuel injector is less than or equal to the static threshold value, updating the current dynamic threshold value according to the similarity of the current operation state data of the fuel injector.

[0034] In some embodiments, the determining whether the operation state of the fuel injector fails according to the time sequence of the current operation state data of the fuel injector, comprises:

[0035] determining the number of times that the current operation state data of the fuel injector continuously appears abnormal according to the time sequence of the current operation state data of the fuel injector;

[0036] determine whether the number of times of continuous abnormality of the current operating state data of the fuel injector exceeds a preset number, and if the number exceeds the preset number, determine that a fault occurs in the operating state of the fuel injector and send an alarm information.

[0037] In some embodiments, the similarity of the current operating state data of the fuel injector is determined according to the current operating state data and the fuel injector state estimation model based on the multi-state estimation technology, including:

[0038] The current operating state data is input into the fuel injector state estimation model based on the multi-state estimation technology to obtain an estimated value of the current operating state data of the fuel injector.

[0039] The similarity of the current operating state data of the fuel injector is calculated according to the estimated value of the current operating state data of the fuel injector and the current operating state data.

[0040] In some embodiments, the current operating state data at least includes one of current rail pressure data of the fuel injector, current oil temperature data of the fuel injector and current oil quantity data of the fuel injector.

[0041] The historical operating state data at least includes one of historical rail pressure data of the fuel injector, historical oil temperature data of the fuel injector and historical oil quantity data of the fuel injector.

[0042] Correspondingly, the fuel injector fault early warning device provided by the embodiments of the present application includes:

[0043] The first acquisition module is configured to acquire the current operating state data and the historical operating state data of the fuel injector.

[0044] The model construction module is configured to construct a fuel injector state estimation model based on a multi-state estimation technology according to the historical operating state data.

[0045] The steady state condition judgment module is configured to determine whether the current operating condition of the fuel injector is a steady state condition according to the current operating state data.

[0046] The similarity determination module is configured to determine the similarity of the current operating state data of the fuel injector according to the current operating state data and the fuel injector state estimation model based on the multi-state estimation technology.

[0047] The second acquisition module is configured to acquire a current dynamic threshold.

[0048] The fault judgment module is configured to determine whether a fault occurs in the operating state of the fuel injector according to the current dynamic threshold and the similarity of the current operating state data of the fuel injector.

[0049] The threshold value updating judgment module is configured to judge whether to update the current dynamic threshold value.

[0050] Correspondingly, the electronic device provided in the embodiments of the present application comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the fuel injector fault early warning method when executing the computer program.

[0051] Correspondingly, the computer readable storage medium provided in the embodiments of the present application has a computer program stored thereon, and the computer program is executed by a processor to implement the fuel injector fault early warning method.

[0052] Advantages: Compared with the prior art, the fuel injector fault early warning method, device, equipment and storage medium provided in the embodiments of the present application, the fuel injector fault early warning method comprises: acquiring current running state data and historical running state data of a fuel injector, and constructing a fuel injector state estimation model based on a multi-state estimation technology according to the historical running state data; determining whether the current running state data of the fuel injector is a steady state condition according to the current running state data, and if so, determining the similarity of the current running state data of the fuel injector according to the current running state data and the fuel injector state estimation model based on the multi-state estimation technology; acquiring a current dynamic threshold value, and determining whether the running state of the fuel injector has failed according to the current dynamic threshold value and the similarity of the current running state data of the fuel injector, and determining whether to update the current dynamic threshold value. Therefore, the method can realize early warning before the running state of the fuel injector fails, thereby improving the equipment maintenance efficiency, improving the equipment reliability, and reducing the equipment maintenance cost. Moreover, the state estimation model is constructed based on the historical running state data of the fuel injector, the similarity of the current running state data is calculated based on the state estimation model, and the running state of the fuel injector is determined in real time in combination with the dynamic threshold value, and the combination of the state estimation model and the dynamic threshold value can improve the accuracy, effectiveness and reliability of the fault early warning judgment. BRIEF DESCRIPTION OF DRAWINGS

[0053] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0054] Figure 1 is a structure schematic diagram of a direct current power transmission system of the embodiments of the present application;

[0055] Figure 2 is a whole flow process schematic diagram of a power transmission system receiving end AC fault ride through method of the embodiments of the present application;

[0056] Figure 3 This is a schematic diagram illustrating the construction process of an injector state estimation model provided in an embodiment of this application;

[0057] Figure 4 This is a flowchart illustrating a fuel injector fault early warning method provided in an embodiment of this application;

[0058] Figure 5 This is a schematic diagram of the principle structure of an injector fault early warning device provided in the embodiments of this application;

[0059] Figure 6 This is a structural diagram of an electronic device provided in the embodiments of this application.

[0060] Figure label:

[0061] 101-First acquisition module; 102-Model construction module; 103-Steady-state condition judgment module; 104-Similarity determination module; 105-Second acquisition module; 106-Fault judgment module; 107-Threshold update judgment module; 100-Injector fault early warning device. Detailed Implementation

[0062] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0063] It should be understood that although the terms first, second, etc., may be used herein to describe various components, these components should not be limited by these terms. These terms are used to distinguish one component from another. Therefore, the first component discussed below may be referred to as the second component without departing from the teachings of this application. As used herein, the term "and / or" includes all combinations of any and more of the associated listed items.

[0064] Those skilled in the art will understand that the accompanying drawings are merely schematic diagrams of exemplary embodiments and may not be to scale. The modules or processes shown in the drawings are not necessarily essential for implementing this application and therefore should not be used to limit the scope of protection of this application.

[0065] Figure 1is a flow chart of a fuel injector fault early warning method provided in the embodiments of the present application. The method is applicable to a ship diesel engine fault diagnosis platform to realize the process of early warning of the fault of the fuel injector. The method can be executed by a fuel injector fault early warning device, which can be realized by software and / or hardware, and the device can be configured in the processor of the ship diesel engine fault diagnosis platform. Please refer to Figure 1 The fuel injector fault early warning method comprises the following steps:

[0066] In step 110, current running state data and historical running state data of the fuel injector are acquired, and a Multivariate State Estimation Technique (MSET) based fuel injector state estimation model is constructed according to the historical running state data.

[0067] The fuel injector can be an electronic control fuel injector, and the specific type can be set according to the actual situation, which is not limited here.

[0068] The current running state data can be acquired in real time by a running state monitoring device of the fuel injector, such as a sensor. The historical running state data can be directly acquired from a database or a memory, and the historical monitoring running state data can be directly used without classification (such as marking or labeling), so that the data availability is high.

[0069] The running state data of the fuel injector can be rail pressure data, oil temperature data and oil quantity data, and the specific type can be set according to the actual situation, which is not limited here.

[0070] The establishment of the MSET based fuel injector state estimation model is helpful to the subsequent accurate prediction of the similarity of the current running state data of the fuel injector, and thus is conducive to improving the accuracy and reliability of the subsequent fault prediction.

[0071] As an implementation mode, the current running state data at least includes one of the current rail pressure data of the fuel injector, the current oil temperature data of the fuel injector and the current oil quantity data of the fuel injector; and the historical running state data at least includes one of the historical rail pressure data of the fuel injector, the historical oil temperature data of the fuel injector and the historical oil quantity data of the fuel injector.

[0072] It should be noted that the specific data of the running state data of the fuel injector can be set according to the actual situation, which is not limited here.

[0073] In step 120, it is judged according to the current running state data whether the current running condition of the fuel injector is a steady state condition, and if so, the similarity of the current running state data of the fuel injector is determined according to the current running state data and the fuel injector state estimation model based on the MSET technology.

[0074] The similarity of the current running state data of the fuel injector is related to specific running state data. For example, assuming that the current running state data is the current oil amount data of the fuel injector, the similarity of the current oil amount data of the fuel injector can be determined according to the current oil amount data of the fuel injector and the fuel injector state estimation model based on the MSET. Assuming that the current running state data is the current oil temperature data of the fuel injector, the similarity of the current oil temperature data of the fuel injector can be determined according to the current oil temperature data of the fuel injector and the fuel injector state estimation model based on the MSET.

[0075] Specifically, after the current running state data of the fuel injector is obtained, it is judged according to the current running state data whether the current running condition of the fuel injector is a steady state condition. If it is a steady state condition, the similarity of the current running state data of the fuel injector is further calculated for subsequent fault judgment.

[0076] In step 130, the current dynamic threshold is obtained, and it is judged according to the current dynamic threshold and the similarity of the current running state data of the fuel injector whether the running state of the fuel injector has a fault, and whether the current dynamic threshold is updated.

[0077] In the technical scheme of the embodiment, the working principle of the fuel injector fault early warning method is as follows: Figure 1, first, the current operating state data and the historical operating state data of the fuel injector are acquired, and a fuel injector state estimation model based on a multivariate state estimation technique is constructed according to the historical operating state data. Then, whether the current operating condition of the fuel injector is a steady-state condition is judged according to the current operating state data, and if so, the similarity of the current operating state data of the fuel injector is determined according to the current operating state data and the fuel injector state estimation model based on the multivariate state estimation technique. Finally, the current dynamic threshold is acquired, and whether the operating state of the fuel injector has failed is judged according to the current dynamic threshold and the similarity of the current operating state data of the fuel injector, and whether the current dynamic threshold is updated is judged. Therefore, the fuel injector state estimation model based on MSET is constructed through the historical operating state data of the fuel injector, the similarity of the current operating state data of the fuel injector is determined according to the fuel injector state estimation model based on MSET, and the operating state of the fuel injector can be real-time online fault early warning according to the similarity and the current dynamic threshold, so as to predict whether the fuel injector has failed in advance, improve the equipment maintenance efficiency and reliability. And by combining the fuel injector state estimation model with the current dynamic threshold acquired in real time, real-time online early warning of fuel injector failure can be realized, and by updating the current dynamic threshold in real time, the accuracy, effectiveness and reliability of the early warning can be improved. In addition, by establishing the fuel injector state estimation model according to the historical operating state data of the fuel injector, on the one hand, the similarity of the current operating state data of the fuel injector can be accurately predicted through the fuel injector state estimation model and the current operating state data of the fuel injector, so as to improve the accuracy of subsequent fault prediction. On the other hand, without classification and other processing of historical data, the historical data can be directly used for construction of the fuel injector state estimation model, the data availability is high, and the problem of low data availability caused by complex long-period monitoring data conditions and no label is solved.

[0078] The technical scheme of the embodiment provides a fuel injector fault early warning method, which comprises the following steps: acquiring current operation state data and historical operation state data of a fuel injector, and constructing a fuel injector state estimation model based on a multivariate state estimation technology according to the historical operation state data; determining whether the current operation state of the fuel injector is a steady state condition according to the current operation state data, and if yes, determining the similarity of the current operation state data of the fuel injector according to the current operation state data and the fuel injector state estimation model based on the multivariate state estimation technology; acquiring a current dynamic threshold, and determining whether the operation state of the fuel injector is faulty according to the current dynamic threshold and the similarity of the current operation state data of the fuel injector, and determining whether to update the current dynamic threshold. Therefore, the method can realize early warning before the operation state of the fuel injector is faulty, thereby improving equipment maintenance efficiency, improving equipment reliability, and reducing equipment maintenance cost. In addition, the state estimation model is constructed according to the historical operation state data of the fuel injector, the similarity of the current operation state data is calculated based on the state estimation model, and whether the operation state of the fuel injector is faulty is determined in real time in combination with the dynamic threshold, so that the combination of the state estimation model and the dynamic threshold can improve the accuracy, effectiveness and reliability of fault early warning and determination.

[0079] Figure 2 is a flowchart of another fuel injector fault early warning method provided in the embodiment. On the basis of the above-mentioned embodiment, optionally, referring to Figure 2 , the method comprises the following steps:

[0080] Step 210: acquiring current operation state data and historical operation state data of a fuel injector, and screening historical steady state condition data according to the historical operation state data.

[0081] As an implementation manner, optionally, screening the historical steady state condition data according to the historical operation state data comprises: determining whether the historical operation state data is a steady state condition according to the historical operation state data, eliminating non-steady state condition data in the historical operation state data, and retaining steady state condition data.

[0082] The non-steady state condition data is variable condition stage data. The specific variable condition type can be set according to actual conditions, and is not limited here.

[0083] Step 220: establishing a fuel injector state estimation model based on a multivariate state estimation technology according to the historical steady state condition data.

[0084] As an implementation, optionally, the injector state estimation model based on the multivariate state estimation technology is established according to historical steady state operating condition data, including: pre-processing the historical steady state operating condition data to obtain a test matrix and a training matrix; generating a model observation matrix based on the training matrix; and establishing the injector state estimation model based on the multivariate state estimation technology according to the test matrix and the model observation matrix.

[0085] The pre-processing of the historical steady state operating condition data includes: first, removing data missing points and outliers in the historical steady state operating condition data (for example, judging whether there are parameters out of range and parameters far exceeding the alarm value in the data, and if so, removing them as outlier samples), and then performing standardization processing. The data after standardization processing is used as sample data. A certain proportion of sample data is randomly selected as training data, and the rest is used as test data. The training matrix is obtained according to the training data, and the test matrix is obtained according to the test data.

[0086] As an implementation, optionally, the model observation matrix is generated based on the training matrix, including: determining a health matrix according to the training matrix; calculating a model nonlinear operator based on the Euclidean distance; determining a model weight coefficient according to the nonlinear operator and the health matrix; and generating the model observation matrix according to the model weight coefficient.

[0087] The specific implementation of determining the model weight coefficient according to the nonlinear operator and the health matrix can be: introducing a system relative error calculation formula, and combining the nonlinear operator and the health matrix to obtain the model weight coefficient.

[0088] Step 230, judging whether the current operating condition of the injector is a steady state operating condition according to the current operating state data, if yes, inputting the current operating state data into the injector state estimation model based on the multivariate state estimation technology to obtain an estimated value of the current operating state data of the injector.

[0089] After it is judged that the current operating condition is a steady state operating condition according to the current operating state data according to the judgment method of the steady state operating condition, the subsequent fault judgment is performed, which is beneficial to improve the accuracy of the subsequent judgment, thereby improving the accuracy, effectiveness and reliability of the fault warning.

[0090] Step 240, calculating the similarity of the current operating state data of the injector according to the estimated value and the current operating state data of the current operating state data of the injector.

[0091] For example, assuming that the current operating state data of the fuel injector is the oil temperature of the fuel injector, the oil temperature of the fuel injector is input into the MSET-based fuel injector state estimation model to obtain an estimated value (or model predicted value) of the oil temperature of the fuel injector, and then the estimated value of the oil temperature of the fuel injector and the oil temperature of the fuel injector (i.e. the actual value) are input into the similarity calculation formula to obtain the similarity of the oil temperature of the fuel injector. Similarly, assuming that the current operating state data of the fuel injector is the oil amount of the fuel injector, the oil amount of the fuel injector is input into the MSET-based fuel injector state estimation model to obtain an estimated value (or model predicted value) of the oil amount of the fuel injector, and then the estimated value of the oil amount of the fuel injector and the oil amount of the fuel injector are input into the similarity calculation formula to obtain the similarity of the oil amount of the fuel injector.

[0092] The similarity calculation formula of the current operating state data of the fuel injector is:

[0093]

[0094] wherein x i is the actual value of the i-th data point of the current operating state data of the fuel injector, y i is the model predicted value (i.e. the estimated value) of the i-th data point of the current operating state data of the fuel injector, and n is a natural number.

[0095] Step 250, judge whether the similarity of the current operating state data of the fuel injector is greater than the current dynamic threshold value. If yes, execute step 260; otherwise, execute step 270.

[0096] wherein different operating state data corresponds to different dynamic threshold values. For example, the current dynamic threshold value corresponding to the current operating state data of the oil temperature of the fuel injector is different from the current dynamic threshold value corresponding to the current operating state data of the oil amount of the fuel injector.

[0097] Specifically, the similarity of the current operating state data of the oil injector is compared with the current dynamic threshold value, and it is determined whether the current operating state data is abnormal, and then it is determined whether the current operating state of the oil injector is faulty. For example, when the current operating state data of the oil injector is the oil temperature of the oil injector, a sliding window length value (the specific value can be set according to the actual situation, which is not specifically limited here) is set to collect a period of oil temperature data of the oil injector, and then the oil temperature of the oil injector is input into the oil injector state estimation model based on MSET, and the estimated value of the oil temperature of the oil injector can be calculated. Then, according to the estimated value of the oil temperature of the oil injector and the oil temperature of the oil injector, the similarity of the oil temperature of the oil injector can be calculated, and then the similarity of the oil temperature of the oil injector is compared with the corresponding current dynamic threshold value to determine whether the oil temperature of the oil injector is faulty. In order to predict whether the oil temperature is faulty in advance and give an alarm, thereby improving the equipment maintenance efficiency, improving the equipment reliability, and reducing the equipment maintenance cost. By combining the oil injector state estimation model with the current dynamic threshold value of the oil temperature of the oil injector obtained in real time, real-time online early warning of oil temperature fault of the oil injector can be realized, and by updating the current dynamic threshold value corresponding to the oil temperature of the oil injector in real time, the accuracy, effectiveness and reliability of the early warning can be improved.

[0098] Similarly, when the current operating state data of the oil injector is the oil quantity of the oil injector, the above method can be used to determine whether the oil quantity of the oil injector is faulty, so as to predict whether the oil quantity is faulty in advance and give an alarm, thereby improving the equipment maintenance efficiency, improving the equipment reliability, and reducing the equipment maintenance cost. By combining the oil injector state estimation model with the current dynamic threshold value of the oil quantity of the oil injector obtained in real time, real-time online early warning of oil quantity fault of the oil injector can be realized, and by updating the current dynamic threshold value corresponding to the oil quantity of the oil injector in real time, the accuracy, effectiveness and reliability of the early warning can be improved.

[0099] Step 260, determining that the current operating state data of the oil injector is not abnormal, and determining whether to update the current dynamic threshold value.

[0100] Specifically, the similarity of the current operating state data of the oil injector is compared with the current dynamic threshold value, and if the similarity of the current operating state data of the oil injector is greater than the current dynamic threshold value, it can be determined that the current operating state data of the oil injector is not abnormal, and further, it is determined whether the current dynamic threshold value needs to be updated to improve the accuracy, effectiveness and reliability of the fault early warning.

[0101] As an implementation, optionally, the judging whether to update the current dynamic threshold value comprises: comparing the similarity of the current operating state data of the fuel injector with the static threshold value; if the similarity of the current operating state data of the fuel injector is greater than the static threshold value, maintaining the current dynamic threshold value; if the similarity of the current operating state data of the fuel injector is less than or equal to the static threshold value, updating the current dynamic threshold value according to the similarity of the current operating state data of the fuel injector.

[0102] The initial dynamic threshold value of the current dynamic threshold value can be set as the initial similarity of each interval. For example, after judging that the current operating condition of the fuel injector is a steady state condition, inputting the current operating state data of the fuel injector collected in real time into the MSET-based fuel injector state estimation model, calculating the similarity of the current operating state data, setting the sliding window length value, and obtaining the initial similarity calculated by the MSET-based fuel injector state estimation model as the initial dynamic threshold value of the current dynamic threshold value.

[0103] The static threshold value is greater than the initial dynamic threshold value of the current dynamic threshold value.

[0104] The setting method of the current dynamic threshold value can be: adjusting the dynamic threshold value at the previous sampling time according to the mean and variance of the similarity at the current sampling time and the mean and variance of the similarity at the previous sampling time.

[0105] Specifically, when the similarity of the current operating state data of the fuel injector is greater than the current dynamic threshold value, it indicates that the current operating state data is normal and the fuel injector is also normal, and then the similarity of the current operating state data of the fuel injector (i.e. the operating state data collected at the current sampling time) is compared with the static threshold value to judge whether to update the current dynamic threshold value. If the similarity of the current operating state data of the fuel injector is greater than the static threshold value, the current dynamic threshold value is maintained. If the similarity of the current operating state data of the fuel injector is less than or equal to the static threshold value, the similarity of the current operating state data of the fuel injector is taken as the current dynamic threshold value, and the current dynamic threshold value is updated. Thus, after judging whether the current operating state of the fuel injector is abnormal at each sampling time, if the current operating state is normal, the current dynamic threshold value is updated, and by updating the current dynamic threshold value in real time, the accuracy of fault prediction can be improved.

[0106] Step 270, judging whether the current operating state data of the fuel injector is abnormal, and obtaining a time sequence of the result of judging whether the current operating state data of the fuel injector is abnormal; judging whether the operating state of the fuel injector is faulty according to the time sequence of the result of judging whether the current operating state data of the fuel injector is abnormal.

[0107] The time sequence of the current operation state data abnormality judgment result of the fuel injector refers to a judgment result obtained by judging whether the current operation state data of the fuel injector is abnormal within a period of time (or within a certain sampling time).

[0108] As an implementation, optionally, judging whether the operation state of the fuel injector is faulty according to the time sequence of the current operation state data abnormality judgment result of the fuel injector includes: determining the number of times that the current operation state data of the fuel injector is continuously abnormal according to the time sequence of the current operation state data abnormality judgment result of the fuel injector; and judging whether the number of times that the current operation state data of the fuel injector is continuously abnormal exceeds a preset number of times, if yes, determining that the operation state of the fuel injector is faulty, and sending an alarm information.

[0109] The specific value of the preset number of times can be set according to actual conditions, and is not specifically limited here.

[0110] Specifically, the similarity of the current operation state data of the fuel injector is compared with the current dynamic threshold value, if the similarity of the current operation state data of the fuel injector is less than or equal to the current dynamic threshold value, it can be judged that the current operation state data of the fuel injector is possibly abnormal. In order to improve the accuracy of the judgment, further, the time sequence of the current operation state data abnormality judgment result of the fuel injector is obtained, and the number of times that the current operation state data is abnormal can be determined according to the time sequence of the judgment result. Then, the number of times that the current operation state data is abnormal is compared with a preset number of times, if greater than the preset number of times, it can be judged that the operation state of the fuel injector is possibly faulty, and an alarm information is sent. Thus, the operation state of the fuel injector can be real-time online early warned, so that the fault can be predicted in advance before the fuel injector is faulty, and thus the equipment maintenance cost can be reduced, and the equipment maintenance efficiency and reliability can be improved.

[0111] In the technical scheme of the embodiment, the working principle of the fuel injector fault early warning method is as follows: Figure 2, first, the current operating state data and the historical operating state data of the fuel injector are acquired, and the historical steady state operating condition data is screened according to the historical operating state data. Then, the fuel injector state estimation model based on the multivariate state estimation technology is established according to the historical steady state operating condition data. Secondly, whether the current operating condition of the fuel injector is a steady state operating condition is judged according to the current operating state data. If yes, the current operating state data is input into the fuel injector state estimation model based on the multivariate state estimation technology to obtain the estimated value of the current operating state data of the fuel injector. The similarity of the current operating state data of the fuel injector is calculated according to the estimated value and the current operating state data of the current operating state data of the fuel injector. Whether the similarity of the current operating state data of the fuel injector is greater than the current dynamic threshold is judged. If the similarity of the current operating state data of the fuel injector is greater than the current dynamic threshold, it is determined that the current operating state data of the fuel injector is normal, and whether to update the current dynamic threshold is further judged. If the similarity of the current operating state data of the fuel injector is less than or equal to the current dynamic threshold, it is determined that the current operating state data of the fuel injector is abnormal, and the time sequence of the current operating state data of the fuel injector is acquired. Whether the operating state of the fuel injector fails is judged according to the time sequence of the current operating state data of the fuel injector. Therefore, the fuel injector state estimation model based on MSET is constructed by using the historical operating state data of the fuel injector, the similarity of the current operating state data of the fuel injector is determined according to the fuel injector state estimation model based on MSET, and the operating state of the fuel injector can be real-time online fault early warning according to the similarity and the current dynamic threshold, so as to predict whether the fuel injector fails in advance, improve the equipment maintenance efficiency and reliability. And by combining the fuel injector state estimation model with the real-time acquired current dynamic threshold, the real-time online early warning of the fuel injector failure can be realized, and by updating the current dynamic threshold in real time, the accuracy, effectiveness and reliability of the early warning can be improved. In addition, by establishing the fuel injector state estimation model according to the historical operating state data of the fuel injector, on the one hand, the similarity of the current operating state data of the fuel injector can be accurately predicted by the fuel injector state estimation model and the current operating state data of the fuel injector, so as to improve the accuracy of subsequent fault prediction. On the other hand, without classification and other processing of historical data, the historical data can be directly used for construction of the fuel injector state estimation model, the data availability is high, and the problem of low data availability caused by complex long-period monitoring data and no label is solved.

[0112] As a specific implementation, optionally, the judgment of the steady state operating condition comprises: acquiring the operating state data of the fuel injector; wherein the operating state data comprises the oil temperature data of the fuel injector and the rail pressure data of the fuel injector; judging whether the rail pressure of the fuel injector changes according to the rail pressure data of the fuel injector, if yes, further judging whether the monotonicity of the oil temperature of the fuel injector changes, if yes, determining that the current operating condition is a steady state operating condition.

[0113] Specifically, first, it is judged whether the rail pressure of the fuel injector changes. If the rail pressure changes, it is indicated that the current data is the variable working condition data. Then, it is further judged whether the monotonicity of the oil temperature data changes. If the monotonicity of the oil temperature changes, it is indicated that the temperature under the current variable working condition changes from the rising or falling trend to the stable (or constant value), and it can be determined that the current working condition is the steady state working condition. Thus, the accuracy of the steady state working condition judgment can be improved, and the inaccurate judgment can be avoided, so that the subsequent fault early warning judgment is not inaccurate.

[0114] For example, in the process of online early warning of the fault of the fuel injector (such as when the fuel injector is assembled and used), according to the current running state data, the steady state working condition judgment method can be used to judge whether the current running working condition of the fuel injector is the steady state working condition, so as to make the subsequent fault early warning judgment. For example, assuming that the current running state data is the rail pressure data and the oil temperature data of the fuel injector, the steady state working condition judgment method can be used to judge whether the current running working condition of the fuel injector is the steady state working condition.

[0115] For example, in the process of constructing the fuel injector state estimation model based on MSET, according to the historical running state data, the steady state working condition judgment method can be used to judge whether the historical running working condition of the fuel injector is the steady state working condition, so as to filter out the historical steady state working condition data, so as to make the subsequent modeling. For example, assuming that the historical running state data is the rail pressure data and the oil temperature data of the long-period fuel injector, the steady state working condition judgment method can be used to judge whether the current running working condition of the fuel injector is the steady state working condition.

[0116] Figure 3 FIG. 1 is a flowchart of a method for constructing a fuel injector state estimation model according to an embodiment of the present application. As shown in FIG. 1, the method comprises the following steps. Figure 3, obtain historical running state data (such as long-period rail pressure and injector temperature monitoring data), determine whether the rail pressure changes, if the rail pressure changes, determine that the current data is variable working condition data, and further determine whether the monotonicity of the temperature of the injector changes. If the monotonicity of the temperature of the injector changes, it is determined that the current data is steady-state working condition data, and non-steady-state working condition data (i.e. variable working condition data) is removed, and steady-state working condition data is retained. Then, remove data missing points and outliers. Secondly, standardize the data. After standardization, a certain proportion of samples are randomly selected as training data, and the rest are test data. Then, the parameters are screened to determine the training matrix K. The health matrix D is calculated based on the training matrix K. The system nonlinear operator is calculated based on the Euclidean distance. The model weight coefficient W is calculated based on the system relative error. Then, the model observation matrix is generated. The relative error of the model observation matrix to the test matrix is calculated, and the relative error distribution is calculated. Then, it is determined whether the equal error meets the requirements, if it meets the requirements, the construction of the injector state estimation model based on MSET is completed, otherwise, return to the step of screening parameters to determine the training matrix K and repeat the execution until the requirements are met.

[0117] Exemplarily, taking the historical rail pressure data of the injector and the historical oil temperature data of the injector as examples, the construction process of the injector state estimation model based on MSET is as follows: first, based on the filtered historical steady-state working condition data, such as historical rail pressure data and historical oil temperature data, remove the data missing point samples. Then, determine whether there are parameters that exceed the range and parameters that are far beyond the alarm value in the data, if so, remove them as outliers to obtain sample data. Secondly, the sample data is converted into dimensionless values according to the data conversion formula. Exemplarily, the value range can be [0, 1]. The data conversion formula is:

[0118]

[0119] wherein, x z-score is the converted dimensionless value result; x i is the i-th data point; x min is the minimum value in the data set; x max is the maximum value in the data set.

[0120] Secondly, the standardized data of the historical steady-state working condition after removing the missing points and outliers is taken as the analysis object, and is divided into training data and test data according to a certain proportion (the specific proportion setting can be set according to the actual situation, which is not limited here).

[0121] Secondly, the training matrix K is determined as:

[0122] K = [X (t1) , X (t2)X (t3) ,…,X (tk) ]

[0123] Secondly, the health matrix D is calculated according to the health matrix calculation formula. The health matrix calculation formula is:

[0124]

[0125] Secondly, the nonlinear operator is calculated based on the Euclidean distance The calculation formula of the nonlinear operator is:

[0126]

[0127] Wherein, x i and y i are actual value and predicted value respectively.

[0128] Secondly, the system relative error calculation formula is introduced:

[0129] Epsilon = |X obs -X est | = |X obs -D·W|

[0130] Wherein, X obs is the system test matrix; X est is the system estimation vector.

[0131] Secondly, the partial derivative of W = [w1, w2, …, w m ] in the system relative error formula is obtained according to the following formula:

[0132] D T ·D·W = D T ·X obs

[0133] Then the matrix operation is carried out, and the following is obtained:

[0134] W = (D T ·D) -1 · (D T ·X obs )

[0135] The model weight coefficient W can be obtained by combining the nonlinear operator. The model weight coefficient calculation formula is:

[0136]

[0137] Finally, the prediction result of the MSET model is compared with the test data, and the relative error between the two is calculated by the following error formula. When the relative error meets the requirement, the fuel injector state estimation model based on MSET is obtained, and if the relative error does not meet the requirement, the step of determining the training matrix is returned and repeated until the relative error meets the requirement.

[0138] The relative error calculation formula is as follows:

[0139]

[0140] wherein δ i is the relative error, x i is the model calculation result, is the actual result of the test data.

[0141] Figure 4 is a flowchart of a fuel injector fault early warning method provided in an embodiment of the present application. For example, refer to Figure 4 When the fuel injector is assembled and used, the health matrix of the fuel injector, the model weight coefficient, and the existing fuel injector temperature and rail pressure steady state working condition data are stored in the control system host computer. Real-time electronic control fuel injector oil temperature and rail pressure data are collected. According to the above steady state working condition judgment method, it is judged whether the current is in a steady state working condition, and the fuel injector fault real-time prediction is executed when it is judged to be in a steady state working condition. The real-time collected fuel injector oil temperature and oil quantity data are input into the fuel injector state estimation model based on MSET, and the similarity between the model estimation value and the actual value is calculated according to the similarity calculation formula. Then, the sliding window length value is selected, and the initial dynamic threshold and static threshold of the model are set. Secondly, the relationship between the similarity and the dynamic threshold is judged, and when the similarity is less than the current dynamic threshold, it is judged that the fuel injector may have an abnormality. When the similarity continues to be less than the dynamic threshold, it is judged that the fuel injector may have an abnormality, and an alarm information is sent. Then, the real-time collected fuel injector oil quantity, oil temperature and rail pressure steady state working condition data are stored in the control system host computer. Thus, the fuel injector fault is realized for real-time online early warning.

[0142] Correspondingly, the present application also provides a fuel injector fault early warning device. Figure 5 is a principle structure block diagram of a fuel injector fault early warning device provided in an embodiment of the present application. Please refer to Figure 5The injector fault early warning device 100 includes: a first acquisition module 101, used to acquire current operating status data and historical operating status data of the injector; a model building module 102, used to construct an injector state estimation model based on multivariate state estimation technology based on the historical operating status data; a steady-state condition judgment module 103, used to judge whether the current operating condition of the injector is a steady-state condition based on the current operating status data; a similarity determination module 104, used to determine the similarity between the current operating status data and the injector state estimation model based on multivariate state estimation technology; a second acquisition module 105, used to acquire a current dynamic threshold; a fault judgment module 106, used to judge whether the operating state of the injector has malfunctioned based on the similarity between the current dynamic threshold and the current operating status data of the injector; and a threshold update judgment module 107, used to judge whether to update the current dynamic threshold.

[0143] The technical solution of this embodiment provides a fuel injector fault early warning device, which includes: a first acquisition module for acquiring current operating status data and historical operating status data of the fuel injector; a model building module for constructing a fuel injector state estimation model based on multivariate state estimation technology based on the historical operating status data; a steady-state condition judgment module for judging whether the current operating condition of the fuel injector is a steady-state condition based on the current operating status data; a similarity determination module for determining the similarity between the current operating status data and the fuel injector state estimation model based on multivariate state estimation technology; a second acquisition module for acquiring a current dynamic threshold; a fault judgment module for judging whether the operating state of the fuel injector has failed based on the similarity between the current dynamic threshold and the current operating status data of the fuel injector; and a threshold update judgment module for judging whether to update the current dynamic threshold. Therefore, this device can provide early warning before a fuel injector fails, thereby improving equipment maintenance efficiency, increasing equipment reliability, and reducing equipment maintenance costs. Furthermore, a state estimation model is constructed using historical operating status data of the injectors. Based on the state estimation model, the similarity of the current operating status data is calculated, and dynamic thresholds are combined to determine in real time whether the operating status of the injectors has failed. The combination of the state estimation model and dynamic thresholds can improve the accuracy, effectiveness, and reliability of fault early warning judgment.

[0144] In some embodiments, the model building module 102 includes: a data filtering unit for filtering historical steady-state operating condition data based on historical operating state data; and a model building unit for building an injector state estimation model based on multivariate state estimation technology based on the historical steady-state operating condition data.

[0145] In some embodiments, the model establishing unit is further configured to preprocess the historical steady-state working condition data to obtain a test matrix and a training matrix; generate a model observation matrix based on the training matrix; and establish an injector state estimation model based on a multivariate state estimation technique according to the test matrix and the model observation matrix.

[0146] In some embodiments, the model establishing unit is further configured to determine a health matrix according to the training matrix; calculate a model nonlinear operator based on a Euclidean distance; determine a model weight coefficient according to the nonlinear operator and the health matrix; and generate the model observation matrix according to the model weight coefficient.

[0147] In some embodiments, the data screening unit is further configured to determine whether the historical running state data is steady-state working condition data according to the historical running state data, eliminate non-steady-state working condition data in the historical running state data, and retain steady-state working condition data.

[0148] In some embodiments, the steady-state working condition determining module 103 is further configured to obtain running state data of the injector; wherein the running state data comprises oil temperature data of the injector and rail pressure data of the injector; determine whether the rail pressure of the injector changes according to the rail pressure data of the injector, and if so, further determine whether the monotonicity of the oil temperature of the injector changes, and if so, determine that the current working condition is a steady-state working condition.

[0149] In some embodiments, the fault determining module 106 comprises: a first comparison unit configured to compare the similarity of the current running state data of the injector with a current dynamic threshold; a fault determining unit configured to determine that the current running state data of the injector is normal if the similarity of the current running state data of the injector is greater than the current dynamic threshold, and determine whether to update the current dynamic threshold; determine that the current running state data of the injector is abnormal if the similarity of the current running state data of the injector is less than or equal to the current dynamic threshold, and obtain a time sequence of the current running state data of the injector; and determine whether the running state of the injector fails according to the time sequence of the current running state data of the injector.

[0150] In some embodiments, the threshold value updating determining module 107 comprises: a second comparison unit configured to compare the similarity of the current running state data of the injector with a static threshold; and a threshold value updating determining unit configured to maintain the current dynamic threshold if the similarity of the current running state data of the injector is greater than the static threshold; and update the current dynamic threshold according to the similarity of the current running state data of the injector if the similarity of the current running state data of the injector is less than or equal to the static threshold.

[0151] In some embodiments, the fault judging unit is further configured to determine the number of times that the current operating state data of the fuel injector is continuously abnormal according to a time sequence of the abnormality judging results of the current operating state data of the fuel injector; and determine whether the number of times that the current operating state data of the fuel injector is continuously abnormal exceeds a preset number of times, and if so, determine that the operating state of the fuel injector is faulty and send an alarm message.

[0152] In some embodiments, the similarity determining module 104 comprises an estimated value obtaining unit configured to input the current operating state data into a fuel injector state estimation model based on a multivariate state estimation technique to obtain an estimated value of the current operating state data of the fuel injector; and a similarity obtaining unit configured to calculate the similarity of the current operating state data of the fuel injector according to the estimated value and the current operating state data of the current operating state data of the fuel injector.

[0153] In some embodiments, the current operating state data at least comprises one of current rail pressure data of the fuel injector, current oil temperature data of the fuel injector and current oil quantity data of the fuel injector; and the historical operating state data at least comprises one of historical rail pressure data of the fuel injector, historical oil temperature data of the fuel injector and historical oil quantity data of the fuel injector.

[0154] Correspondingly, the embodiments of the present application also provide an electronic device, please refer to Figure 6 , Figure 6 a structural diagram of the electronic device of the embodiments of the present application. The electronic device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the fuel injector fault early warning method when executing the computer program. Since the fuel injector fault early warning method is described in detail above, it will not be described here.

[0155] Correspondingly, the embodiments of the present application also provide a computer readable storage medium having a computer program stored thereon, and the computer program is executed by a processor to implement the steps of the fuel injector fault early warning method. Since the fuel injector fault early warning method is described in detail above, it will not be described here.

[0156] In the above embodiments, the description of each embodiment has its own focus, and the parts not described in detail in a certain embodiment can be referred to the relevant description of other embodiments.

[0157] The oil injector fault early warning method, device, equipment and storage medium provided by the embodiments of the present application are described in detail, and specific examples are applied to explain the principles and implementation modes of the present application. The above embodiment description is only used to help understand the technical solutions and core ideas of the present application. Those skilled in the art should understand that the technical solutions recorded in the foregoing embodiments can be modified, or some technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for early warning of fuel injector malfunction, characterized in that, include: Acquire the current operating status data and historical operating status data of the injector, and construct an injector state estimation model based on multivariate state estimation technology based on the historical operating status data; Based on the current operating status data, determine whether the current operating condition of the injector is a steady state. If so, determine the similarity between the current operating status data and the injector state estimation model based on multivariate state estimation technology. Obtain the current dynamic threshold, determine whether the operating status of the injector has malfunctioned based on the similarity between the current dynamic threshold and the current operating status data of the injector, and determine whether to update the current dynamic threshold. The step of determining whether the injector's operating status has malfunctioned based on the similarity between the current dynamic threshold and the injector's current operating status data, and determining whether to update the current dynamic threshold, includes: Compare the similarity of the current operating status data of the injector with the current dynamic threshold; If the similarity of the current operating status data of the injector is greater than the current dynamic threshold, it is determined that the current operating status data of the injector is normal, and it is determined whether to update the current dynamic threshold. If the similarity of the current operating status data of the injector is less than or equal to the current dynamic threshold, it is determined that the current operating status data of the injector is abnormal, and the time series of the judgment result of whether the current operating status data of the injector is abnormal is obtained; the operating status of the injector is determined to be faulty based on the time series of the judgment result of whether the current operating status data of the injector is abnormal.

2. The injector fault early warning method according to claim 1, characterized in that, The step of constructing an injector state estimation model based on multivariate state estimation technology according to the historical operating state data includes: Historical steady-state operating condition data are filtered out based on the historical operating status data; Based on the historical steady-state operating data, an injector state estimation model based on multivariate state estimation technology is established.

3. The injector fault early warning method according to claim 2, characterized in that, The step of establishing an injector state estimation model based on multivariate state estimation technology according to the historical steady-state operating data includes: The historical steady-state operating condition data are preprocessed to obtain the test matrix and the training matrix; Generate the model observation matrix based on the training matrix; The injector state estimation model based on multivariate state estimation technology is established based on the test matrix and the model observation matrix.

4. The injector fault early warning method according to claim 3, characterized in that, The generation of the model observation matrix based on the training matrix includes: Determine the health matrix based on the training matrix; Nonlinear operators based on Euclidean distance calculation model; The model weight coefficients are determined based on the nonlinear operator and the health matrix. The model observation matrix is ​​generated based on the model weight coefficients.

5. The injector fault early warning method according to claim 2, characterized in that, The step of filtering historical steady-state operating condition data based on the historical operating status data includes: Based on the historical operating status data, determine whether the historical operating status data represents a steady-state condition, remove non-steady-state operating condition data from the historical operating status data, and retain the steady-state operating condition data.

6. The injector fault early warning method according to claim 1 or 5, characterized in that, The determination of the steady-state operating condition includes: Acquire the operating status data of the injector; wherein, the operating status data includes the injector's oil temperature data and the injector's rail pressure data; Based on the injector rail pressure data, determine whether the injector rail pressure has changed. If so, further determine whether the injector oil temperature monotonicity has changed. If so, determine that the current operating condition is a steady-state condition.

7. The injector fault early warning method according to claim 1, characterized in that, The step of determining whether to update the current dynamic threshold includes: The similarity of the current operating status data of the injector is compared with a static threshold. If the similarity of the current operating status data of the injector is greater than the static threshold, then the current dynamic threshold is maintained; If the similarity of the current operating status data of the injector is less than or equal to the static threshold, then the current dynamic threshold is updated based on the similarity of the current operating status data of the injector.

8. The injector fault early warning method according to claim 1, characterized in that, The step of determining whether the injector's operating status has malfunctioned based on the time series of the current operating status data of the injector for abnormality includes: The number of consecutive abnormalities in the current operating status data of the injector is determined based on the time series of the judgment results of whether the current operating status data of the injector is abnormal; It also determines whether the number of consecutive abnormalities in the current operating status data of the injector exceeds a preset number. If it does, it determines that the operating status of the injector has malfunctioned and sends an alarm message.

9. The injector fault early warning method according to claim 1, characterized in that, The step of determining the similarity between the current operating state data of the injector and the injector state estimation model based on multivariate state estimation technology includes: The current operating status data is input into the injector state estimation model based on multivariate state estimation technology to obtain an estimated value of the current operating status data of the injector; The similarity between the current operating status data of the injector and the estimated value of the current operating status data is calculated.

10. The injector fault early warning method according to claim 1, characterized in that, The current operating status data includes at least one of the following: the current rail pressure data of the injector, the current oil temperature data of the injector, and the current oil quantity data of the injector; The historical operating status data includes at least one of the following: historical rail pressure data of the injector, historical oil temperature data of the injector, and historical oil quantity data of the injector.

11. A fuel injector malfunction early warning device, characterized in that, include: The first acquisition module is used to acquire the current operating status data and historical operating status data of the fuel injector; The model building module is used to build an injector state estimation model based on multivariate state estimation technology based on the historical operating state data. The steady-state condition judgment module is used to determine whether the current operating condition of the injector is a steady-state condition based on the current operating state data. The similarity determination module is used to determine the similarity between the current operating state data of the injector and the injector state estimation model based on multivariate state estimation technology; The second acquisition module is used to acquire the current dynamic threshold; The fault determination module is used to determine whether the operating status of the injector has malfunctioned based on the similarity between the current dynamic threshold and the current operating status data of the injector. The threshold update determination module is used to determine whether to update the current dynamic threshold; The fault judgment module includes: a first comparison unit, used to compare the similarity of the current operating status data of the injector with the current dynamic threshold; The fault judgment unit is used to determine that the current operating status data of the injector is normal if the similarity of the current operating status data of the injector is greater than the current dynamic threshold, and to determine whether to update the current dynamic threshold; if the similarity of the current operating status data of the injector is less than or equal to the current dynamic threshold, it is determined that the current operating status data of the injector is abnormal, and the time series of the judgment result of whether the current operating status data of the injector is abnormal is obtained; and the operating status of the injector is determined to be faulty based on the time series of the judgment result of whether the current operating status data of the injector is abnormal.

12. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the injector fault early warning method according to any one of claims 1-10.

13. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the injector fault early warning method according to any one of claims 1-10.

Images