Performance analysis method, system and storage medium of linear internal combustion power generation system

Abstract

This invention discloses a performance analysis method, system, and storage medium for a linear internal combustion power generation system. The method includes: collecting operational data from the linear internal combustion power generation system and preprocessing the operational data; dividing the preprocessed operational data into cyclic segments at appropriate locations; combining cyclic segments within the same cycle to establish a sample library; extracting key operational features from the cyclic sample data in the sample library; classifying the cycles based on the extracted key operational features and establishing a sample and key feature library based on the classified data; and calculating the evaluation indicators and statistical characteristics of the cycles based on the samples and key feature library to achieve performance analysis. Through the technical solution of this invention, accurate cyclic segmentation and classification of system data can be achieved, improving the completeness of data processing and the accuracy and completeness of performance evaluation and analysis, and increasing the efficiency of system performance analysis.

Description

Technical Field

[0001] The present invention relates to the technical field of linear internal combustion power generation systems, and in particular to a method for feature extraction and performance analysis of a linear internal combustion power generation system, a system for feature extraction and performance analysis of a linear internal combustion power generation system, and a storage medium. Background Art

[0002] In the field of new energy, linear internal combustion power generation system, as a new form of power, has become a research hotspot for scholars and enterprises. The linear internal combustion power generation system is mainly composed of a linear internal combustion engine and a linear motor. The combustion of the linear internal combustion engine drives the linear motor to generate electricity and realize the output of electrical energy.

[0003] In order to improve the thermodynamic performance and power output efficiency of linear internal combustion power generation systems, the scientific research team needs to conduct a large number of linear internal combustion power generation system tests, expand the applicable dimensions of linear internal combustion power generation systems, and pursue higher thermal efficiency, power generation efficiency and stability. For traditional internal combustion power generation equipment, mature dynamometers and corresponding dynamometer calculation and analysis software provide guarantees for the analysis and calculation of the combustion conditions and indicated power conditions of the power generation system, and can provide strong support for the next step in improving system performance. However, for linear internal combustion power generation systems, traditional analysis methods and means are no longer applicable, which brings huge challenges to the analysis and processing of experimental data of linear internal combustion power generation systems.

[0004] For the stability experiment of linear internal combustion power generation system, the experimental data of a single linear internal combustion power generation system can reach tens of thousands of sets, among which there are many cases of random misfire, incomplete combustion, stall, etc. Faced with massive experimental data, scientific researchers need to spend a lot of time to capture these abnormal cycles in the experimental data, extract features from these abnormal cycle data and store them for subsequent analysis.

[0005] For the performance optimization test of the linear internal combustion power generation system under variable working conditions, the key kinematic characteristics, thermodynamic characteristics and performance of a single cycle are expected to be accurately captured and recorded to facilitate the subsequent classification and correlation analysis of the experimental cycles. Due to the large frequency range and uncertain dead points, the data cannot be segmented by equidistant segmentation or dead point segmentation. At the same time, for a single cycle, due to the effects of combustion impact, motor damping, etc., the linear internal combustion power generation system has unique motion characteristics, with a large speed variation range and dead point position fluctuations in most strokes within a single cycle. This causes trouble for the cycle segmentation of experimental data. It may cause problems such as repeated cycle segmentation, segmentation omissions, and segmentation misjudgments, which makes it difficult to capture key features in subsequent single cycles. This increases the workload of inspection, verification, and key feature extraction after cycle segmentation.

[0006] The complexity of data processing in linear internal combustion power generation systems far exceeds that of traditional internal combustion power generation systems. There is no system performance analysis and calculation software specifically suitable for linear internal combustion power generation systems. At present, in order to reduce the difficulty of experimental data processing, the experimental time has been shortened, which greatly reduces the experimental test dimension. The processing and analysis methods are partially extracted. This extraction method cannot cover all the situations of the experiment, and cannot guarantee that abnormal situations and certain transient situations are covered. At the same time, it is time-consuming, labor-intensive, and inefficient. These problems have limited the operating condition analysis, abnormal cycle processing, and system performance improvement of linear internal combustion systems to a certain extent. Summary of the invention

[0007] In response to the above problems, the present invention provides a performance analysis method, system and storage medium for a linear internal combustion power generation system. The method divides the operating characteristic data of the internal combustion power generation system, extracts key operating characteristics in units of single cycles, and judges the abnormal cycle conditions of the system based on the key operating characteristics and classifies and stores them at the same time. The method calculates specific evaluation indicators and statistical characteristics to implement performance evaluation and analysis of the system. The method can accurately implement cycle division and cycle classification of system data, improve the degree of perfection of data processing and the accuracy and perfection of performance evaluation and analysis, and improve the efficiency of system performance analysis.

[0008] To achieve the above object, the present invention provides a performance analysis method for a linear internal combustion power generation system, comprising:

[0009] Collecting operation data of the linear internal combustion power generation system and preprocessing the operation data;

[0010] For the pre-processed running data, selecting reasonable positions to divide it into loop segment data;

[0011] Combining the loop segment data of the same loop to establish a sample library;

[0012] Extracting key operation features from the cyclic sample data in the sample library;

[0013] Classify and judge the cycles according to the extracted key operating features, and establish a sample and key feature library according to the classification data;

[0014] Based on the samples and key feature library, evaluation indicators and statistical features of the cycle are calculated to achieve performance analysis.

[0015] In the above technical solution, preferably, the specific process of preprocessing the operation data includes:

[0016] Determine whether there is motion speed data in the operation data, and if not, obtain the motion speed data by differentiating the position data and the time data;

[0017] Performing Gaussian filtering on the motion speed data to obtain relatively smooth motion speed data;

[0018] The cylinder pressure data in the operation data is filtered.

[0019] In the above technical solution, preferably, the specific process of selecting a reasonable position to be divided into loop segment data includes:

[0020] Determine whether the speed change range and position change range before and after the preset segmentation point are within the preset threshold range. If not, use a genetic algorithm to find a segmentation point that meets the preset conditions. If so, use the preset segmentation point.

[0021] Dividing the running data into a plurality of cycles according to the characteristics of the segmentation points, and determining whether there are repeated segmentations and missed segmentations;

[0022] If it is determined that repeated segmentation and missed segmentation exist, the operation data is cyclically divided again according to the characteristics of the segmentation points.

[0023] In the above technical solution, preferably, the cycle segment data belonging to the same cycle are divided into multiple cycle segments, which are respectively entered into a cell array for storage to establish a sample library.

[0024] In the above technical solution, preferably, the specific process of extracting key operation features from the cyclic sample data in the sample library includes:

[0025] The ignition timing position is determined according to the characteristics of the cyclic ignition signal, and the cylinder pressure and speed at the ignition timing are recorded;

[0026] According to the cycle sample data, finding and recording the peak cylinder pressure and the corresponding position of the peak cylinder pressure;

[0027] According to the cycle sample data, finding and recording the top dead center and the bottom dead center;

[0028] According to the cycle sample data, finding and recording the expansion point speed and the expansion point cylinder pressure;

[0029] According to the cycle sample data, the intake signal jump time, intake temperature and intake pressure are calculated and recorded.

[0030] In the above technical solution, preferably, the specific process of classifying and judging the cycles according to the extracted key operation features and establishing samples and key feature libraries according to the classification data includes:

[0031] According to whether the difference between the cylinder pressure at the expansion point and the cylinder pressure at the ignition time is less than 0, it is determined whether the current cycle is a misfire cycle, and all data and feature arrays of the current cycle belonging to the misfire cycle are stored in the misfire cycle sample and key feature library;

[0032] According to whether the difference between the top dead center and the stall preset value is less than 0, it is determined whether the current cycle is a stall cycle, and all data and feature arrays of the current cycle belonging to the stall cycle are stored in the stall cycle sample and key feature library;

[0033] According to whether the difference between the expansion point speed and the preset value of the expansion point speed is less than 0, it is judged whether the current cycle is an incomplete combustion situation, and all data and feature arrays of the current cycle belonging to the incomplete combustion situation are stored in the incomplete combustion situation sample and key feature library, and, according to whether the difference between the ratio of the cylinder pressure and the intake pressure at the ignition time and the preset compression ratio is less than 0, it is further judged that the current incomplete combustion situation is a compression ratio insufficient cycle or a quenching cycle, and the compression ratio insufficient cycle and the quenching cycle are stored in the compression ratio insufficient cycle sample and key feature sub-library and the quenching cycle sample and key feature sub-library respectively;

[0034] According to whether the difference between the cylinder pressure at the ignition time and the abnormal cycle monitoring value is less than 0, it is determined whether the current cycle is other abnormal cycles, and all data and feature arrays of the current cycle belonging to other abnormal cycles are stored in other abnormal cycle samples and key feature libraries;

[0035] All data and feature arrays of the current cycle of a normal cycle that does not belong to a misfire cycle, a stall cycle, an incomplete combustion situation, or other abnormal cycles are stored in a normal cycle sample and key feature library.

[0036] In the above technical solution, preferably, the specific process of calculating the evaluation index and statistical characteristics of the cycle according to the sample and key feature library includes:

[0037] For different types of cycles, the cycle frequency, cycle indicated work and cycle thermal efficiency are calculated as evaluation indicators;

[0038] The calculation formula of cycle efficiency Fre is:

[0039]

[0040] Among them, Δt is the time used in the current cycle;

[0041] The calculation formula of cycle indicated work is:

[0042]

[0043] Among them, P mis the cylinder pressure of the mth cycle, A is the piston area in the cylinder, 1 and km are the starting and ending points of the data of the mth cycle;

[0044] For different types of cycles, the average value of top dead center, the average value of bottom dead center, the average value of peak cylinder pressure, the average value of frequency and the average value of indicated work as well as the mean square error of top dead center and the mean square error of peak cylinder pressure are calculated as statistical features.

[0045] The calculation formula for the average value μ is:

[0046]

[0047] Mean square error σ 2 The calculation formula is:

[0048]

[0049] Where n is the number of cycles involved in the average value calculation, c i is the eigenvalue involved in the calculation cycle.

[0050] In the above technical solution, preferably, the performance analysis method of the linear internal combustion power generation system further includes:

[0051] Based on the calculated evaluation indicators and statistical characteristics of the cycle, a statistical graph characterizing the cycle performance is drawn.

[0052] The present invention further provides a performance analysis system for a linear internal combustion power generation system, which uses the performance analysis method for a linear internal combustion power generation system disclosed in any one of the above technical solutions, including:

[0053] A data preprocessing module, used for collecting operation data of the linear internal combustion power generation system and preprocessing the operation data;

[0054] A loop division module is used to select reasonable positions to divide the pre-processed running data into loop segment data;

[0055] A sample building module is used to combine loop segment data of the same loop to build a sample library;

[0056] A feature extraction module, used for extracting key operation features from the cyclic sample data in the sample library;

[0057] A cycle classification module is used to classify and judge the cycle according to the extracted key operation characteristics, and to establish a sample and key feature library according to the classification data;

[0058] The performance analysis module is used to calculate the evaluation index and statistical characteristics of the cycle according to the samples and key feature library to achieve performance analysis.

[0059] The present invention also proposes a storage medium storing a computer program, which, when executed, can implement a performance analysis method for a linear internal combustion power generation system as disclosed in any one of the above technical solutions.

[0060] Compared with the prior art, the beneficial effects of the present invention are as follows: by segmenting the operating characteristic data of the internal combustion power generation system, key operating characteristics are extracted in units of single cycles, and the cycle abnormalities of the system are judged based on the key operating characteristics and classified and stored at the same time, and specific evaluation indicators and statistical characteristics are calculated to realize performance evaluation and analysis of the system, which can accurately realize the cycle division and cycle classification of system data, improve the degree of perfection of data processing and the accuracy and perfection of performance evaluation and analysis, and improve the efficiency of system performance analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 A flow chart of a method for analyzing performance of a linear internal combustion power generation system disclosed in an embodiment of the present invention;

[0062] Figure 2 A schematic diagram of loop division and sample library creation in a linear internal combustion power generation system performance analysis method disclosed in an embodiment of the present invention;

[0063] Figure 3 A schematic diagram of speed difference and filtering of a linear internal combustion power generation system performance analysis method disclosed in an embodiment of the present invention;

[0064] Figure 4 A schematic diagram of cycle division of a linear internal combustion power generation system performance analysis method disclosed in an embodiment of the present invention;

[0065] Figure 5 The number of cycle classifications of the linear internal combustion power generation system performance analysis method disclosed in one embodiment of the present invention is statistically significant;

[0066] Figure 6 A schematic diagram of cycle indicating work of a linear internal combustion power generation system performance analysis method disclosed in an embodiment of the present invention;

[0067] Figure 7 A schematic diagram of cycle thermal efficiency of a linear internal combustion power generation system performance analysis method disclosed in an embodiment of the present invention;

[0068] Figure 8 A schematic diagram of the cycle frequency of a linear internal combustion power generation system performance analysis method disclosed in an embodiment of the present invention;

[0069] Fig. 9 A schematic diagram of modules of a linear internal combustion power generation system performance analysis system disclosed in an embodiment of the present invention.

[0070] In the figure, the corresponding relationship between each component and the reference numeral is as follows:

[0071] 11. Data preprocessing module, 12. Loop partitioning module, 13. Sample establishment module, 14. Feature extraction module, 15. Loop classification module, 16. Performance analysis module. DETAILED DESCRIPTION

[0072] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0073] The present invention is further described in detail below in conjunction with the accompanying drawings:

[0074] like Figure 1 and Figure 2 As shown, a performance analysis method of a linear internal combustion power generation system provided by the present invention includes:

[0075] Collect the operation data of the linear internal combustion power generation system and pre-process the operation data;

[0076] For the pre-processed running data, select reasonable positions to divide it into loop segment data;

[0077] Combining the loop segment data of the same loop to establish a sample library;

[0078] Extract key operation features from the cyclic sample data in the sample library;

[0079] Classify and judge the cycles according to the extracted key operating features, and establish samples and key feature libraries based on the classified data;

[0080] Based on the samples and key feature libraries, the evaluation indicators and statistical characteristics of the cycle are calculated to achieve performance analysis.

[0081] In this embodiment, the operating characteristic data of the internal combustion power generation system is segmented, key operating characteristics are extracted in units of single cycles, and the cycle abnormalities of the system are judged based on the key operating characteristics and classified and stored at the same time, and specific evaluation indicators and statistical characteristics are calculated to implement performance evaluation and analysis of the system. This can accurately implement cycle division and cycle classification of system data, improve the degree of perfection of data processing and the accuracy and perfection of performance evaluation and analysis, and improve the efficiency of system performance analysis.

[0082] Specifically, the present invention analyzes the operating laws of linear internal combustion power generation systems through in-depth mining and analysis of a large number of test data of linear internal combustion power generation systems (including single-cylinder, dual-cylinder, and multi-module types), and provides a reference for technical personnel in related fields; performs data segmentation based on the operating characteristics of the internal combustion power generation system, and stores relevant test data in units of single cycles, so as to facilitate the extraction of single cycles for analysis at any time; extracts key features of each cycle data after segmentation; studies relevant algorithms to quickly and accurately determine the abnormal conditions of the system (misfire, incomplete combustion, stall, etc.), and automatically classifies and stores them; further, to facilitate horizontal comparison of systems, statistical analysis is performed on key characteristics of the system, and analysis methods and tools are provided for linear internal combustion power generation system test data analysis and processing, system abnormality analysis, system variable operating condition analysis, system evaluation, etc.

[0083] This method fills the gap in the performance analysis of linear internal combustion power generation systems. It has a reasonable design and a perfect processing process. It can select appropriate cycle segmentation points, accurately and quickly segment the data of tens of thousands of cycles, and extract the kinematic characteristics and thermodynamic characteristics within the cycle in parallel. It can capture misfires, incomplete combustion, stalls and other situations in the data, classify, store and output the data, and realize the evaluation of specific performance of linear internal combustion power generation systems in terms of thermal efficiency, indicated power and so on.

[0084] Among them, the necessary data required by the present invention are time t, displacement x, cylinder pressure p, system structural parameters cylinder diameter A, ignition position xi, etc. Other data including intake air temperature T, intake air flow m, exhaust gas oxygen content o, etc. are optional inputs. The necessary data can complete the main performance analysis of the internal combustion power generation system, and the optional data can complete a more comprehensive and accurate performance analysis of the internal combustion power generation system.

[0085] Preferably, the specific process of preprocessing the operation data includes:

[0086] Determine whether there is motion speed data in the running data. If not, obtain the motion speed data by differentiating the position data and the time data.

[0087] Perform Gaussian filtering on the motion speed data to obtain smoother motion speed data;

[0088] Filter the cylinder pressure data in the operating data.

[0089] In the above implementation, preferably, for the pre-processed running data, the specific process of selecting a reasonable position to be divided into loop segment data includes:

[0090] Determine whether the speed change range and position change range before and after the preset segmentation point are within the preset threshold range. If not, use a genetic algorithm to find a segmentation point that meets the preset conditions. If so, use the preset segmentation point to proceed to the next step.

[0091] Divide the running data into multiple cycles according to the characteristics of the segmentation point (x=a&&sign(v(x=a))), and determine whether there are repeated segmentations and missed segmentations;

[0092] If it is determined that there are repeated segmentations and missed segmentations, the other data columns in the running data are re-divided cyclically according to the characteristics of the segmentation points.

[0093] Among them, the genetic algorithm (GA) is a computational model of biological evolution that simulates the natural selection and genetic mechanism of Darwin's theory of biological evolution. It is a method of searching for the optimal solution by simulating the natural evolution process. The algorithm uses mathematical methods and computer simulation operations to convert the problem-solving process into a process similar to the crossover and mutation of chromosome genes in biological evolution.

[0094] In the above implementation, preferably, the cycle segment data x(1:km), p(1:km), ... m(1:km) belonging to the same cycle are divided into multiple cycle segments, which are respectively entered into the cell array mycell for storage to establish a sample library.

[0095] In the above implementation, preferably, the key operation characteristics of key operation characteristics and key thermodynamic characteristics are extracted for the cycle sample data in the sample library, wherein the key kinematic characteristics include: ignition time position, ignition time speed, peak cylinder pressure corresponding position, top dead center, bottom dead center and specific expansion point speed, etc., and the key thermodynamic characteristics include: ignition time cylinder pressure, peak cylinder pressure, injection amount, intake air quality, intake air temperature, intake pressure and specific expansion point cylinder pressure, etc. The specific process includes:

[0096] Determine whether all features of the current cycle have been entered into charic. If so, end feature extraction. Otherwise, proceed to the next step of judgment.

[0097] The ignition timing position is determined according to the characteristics of the cyclic ignition signal, specifically: determine whether sign(ig(k)-ig(k-1))=1 is true, if so, record x=ig(k) into the array charic and jump out, otherwise, continue to determine whether sign(ig(k)-ig(k-1)) is equal to 1;

[0098] Record the cylinder pressure P(ig(k)) at the ignition time and the speed v(ig(k)) at the ignition time to the array charic;

[0099] According to the cycle sample data, find and record the peak cylinder pressure max(P) and the corresponding position x(max(P)) of the peak cylinder pressure into the array charic;

[0100] According to the loop sample data, find and record the top dead point max(x) and the bottom dead point min(x) to the array charic;

[0101] According to the cyclic sample data, find the expansion point ex(j) according to x=ig(k)&&v=-sign(v(ig(k))), and record the expansion point velocity v(ex(j)) and the expansion point cylinder pressure P(ex(j)) to the array charic;

[0102] According to the cycle sample data, calculate and record the intake signal jump time Δt(inta) and the intake mass Δt(inta)*minta to the array charic; calculate and record the injection signal jump time Δt(inj) and the injection amount Δt(inj)*mfuel to the array charic; calculate the intake temperature average(Tin) and the intake pressure average(Pin) to the array charic.

[0103] In the above implementation, preferably, the specific process of classifying and judging the cycles according to the extracted key operation features and establishing the sample and key feature library according to the classification data includes:

[0104] Determine whether the difference between the cylinder pressure at the expansion point and the cylinder pressure at the ignition time P(ex(j))-P(ig(k)) is less than 0. If so, determine that the current cycle is a misfire cycle, and store all the data and feature arrays of the current cycle belonging to the misfire cycle into the misfire cycle sample and key feature library misfire. If not, proceed to the next step of determination.

[0105] Determine whether the difference between the top dead center and the stall preset value X, max(x)-X, is less than 0. If so, determine that the current cycle is a stall cycle, and store all data and feature arrays of the current cycle belonging to the stall cycle into the stall cycle sample and key feature library stall. If not, proceed to the next step of determination.

[0106] Determine whether the difference v(ex(j))-V between the expansion point speed and the preset expansion point speed value V is less than 0. If not, proceed to the next step of determination. If so, determine that the current cycle is an inadequate combustion situation, and store all data and feature arrays of the current cycle belonging to the inadequate combustion situation into the inadequate combustion situation sample and key feature library Inadequate. On this basis, further determine whether the difference P(ig(k)) / average(Pin)-CR between the ratio of the cylinder pressure at the ignition moment and the intake pressure and the preset compression ratio CR is less than 0. If so, determine that the current inadequate combustion situation is an insufficient compression ratio cycle, and store all data and feature arrays of the insufficient compression ratio cycle into the insufficient compression ratio cycle sample and key feature sub-library 1compression. Otherwise, determine that the current inadequate combustion situation is a quenching cycle, and store all data and feature arrays of the quenching cycle into the quenching cycle sample and key feature sub-library 2cuxi.

[0107] Determine whether the difference P(ig(k))-5*P(in) between the cylinder pressure at the ignition time and the abnormal cycle monitoring value 5*P(in) is less than 0. If so, determine that the current cycle is other abnormal cycles, and store all data and feature arrays of the current cycle belonging to other abnormal cycles into other abnormal cycle samples and key feature libraries abnormal. If not, proceed to the next step of judgment.

[0108] If the current cycle does not belong to a misfire cycle, a stall cycle, an incomplete combustion situation or other abnormal cycles, the current cycle is determined to be a normal cycle, and all data and feature arrays of the current cycle are stored in the normal cycle sample and key feature library.

[0109] In the above implementation, preferably, the specific process of calculating the evaluation index and statistical characteristics of the cycle according to the sample and key feature library includes:

[0110] For different types of cycles, the cycle frequency, cycle indicated work and cycle thermal efficiency are calculated as evaluation indicators;

[0111] The calculation formula of cycle efficiency Fre is:

[0112]

[0113] Among them, Δt is the time used in the current cycle;

[0114] The calculation formula of the cycle indicated work through the PV diagram area is:

[0115]

[0116] Among them, P m is the cylinder pressure of the mth cycle, A is the piston area in the cylinder, 1 and km are the starting and ending points of the data of the mth cycle;

[0117] For different types of cycles, the average value of top dead center, the average value of bottom dead center, the average value of peak cylinder pressure, the average value of frequency and the average value of indicated work as well as the mean square error of top dead center and the mean square error of peak cylinder pressure are calculated as statistical features.

[0118] The calculation formula for the average value μ is:

[0119]

[0120] Mean square error σ 2 The calculation formula is:

[0121]

[0122] Where n is the number of cycles involved in the average value calculation, c i is the eigenvalue involved in the calculation cycle.

[0123] In the above embodiment, preferably, the performance analysis method of the linear internal combustion power generation system further includes:

[0124] Based on the calculated evaluation indicators and statistical characteristics of the cycle, a statistical graph characterizing the cycle performance is drawn.

[0125] According to the performance analysis method of the linear internal combustion power generation system disclosed in the above-mentioned embodiment, the present invention specifically illustrates the method through the following examples.

[0126] Embodiment 1:

[0127] Data collection and processing: In this embodiment, the time, position, cylinder pressure, intake temperature, intake pressure, injection signal, and ignition signal data of a single-cylinder linear internal combustion power generation system running for 20.5s at a sampling frequency of 20k are collected.

[0128] If the data lacks speed information, obtain the speed data column through difference and filtering. Figure 3 The obtained speed information (excerpt) is shown;

[0129] Circular division: Verify whether the position information meets the conditions at the preset value -3mm&&sign(v(x=-3))=1. If not, use the genetic algorithm to find a reasonable division point, and obtain the final division feature as follows: Figure 4 The position division fragment divides other information in the original data according to the position division condition to create a sample library mycell. In this example, 385 cycles are divided.

[0130] Feature extraction: The key features extracted include key operational features and key thermodynamic features. The key kinematic features include: ignition timing position, ignition timing speed, peak cylinder pressure corresponding position, top dead center, bottom dead center, specific expansion point speed, etc. The key thermodynamic features include: ignition timing cylinder pressure, peak cylinder pressure, injection amount, intake mass, intake temperature, intake pressure, specific expansion point cylinder pressure, forming a [12×385] charic array;

[0131] Cycle classification: Cycle classification includes misfire cycle, stall cycle, incomplete combustion cycle, other abnormal cycles and normal operation cycle. Incomplete combustion cycle includes quench cycle and insufficient compression ratio cycle. 385 cycles are judged, such as Figure 5 As shown, among the 385 cycles, there were 2 misfire cycles, 1 stall cycle, 47 incomplete combustion cycles, 0 other abnormal cycles, and 335 normal operation cycles, including 36 quenching cycles and 11 cycles with insufficient compression ratio.

[0132] Performance evaluation: Figure 6-8 As shown, the evaluation indicators of the cycle include: cycle frequency, cycle indicated work, and cycle thermal efficiency.

[0133] Calculate the statistical characteristics of the cycle: as shown in Table 1 below, including the top dead center average, top dead center mean square error, bottom dead center average, peak cylinder pressure average, peak cylinder pressure mean square error, frequency average, indicated work average, etc.

[0134] Table 1 Cyclic statistical characteristics

[0135]

[0136] The performance analysis of the linear internal combustion power generation system disclosed in the above-mentioned embodiments has strong portability and universality, and is not limited to the performance analysis of the linear internal combustion power generation system of a certain structural form.

[0137] The present invention further provides a performance analysis system for a linear internal combustion power generation system, which uses the performance analysis method for a linear internal combustion power generation system disclosed in any one of the above embodiments, including:

[0138] The data preprocessing module 11 is used to collect the operation data of the linear internal combustion power generation system and preprocess the operation data;

[0139] The loop division module 12 selects reasonable positions to divide the pre-processed running data into loop segment data;

[0140] The sample building module 13 is used to combine the loop segment data of the same loop to build a sample library;

[0141] A feature extraction module 14 is used to extract key operation features from the cyclic sample data in the sample library;

[0142] The cycle classification module 15 is used to classify and judge the cycle according to the extracted key operation characteristics, and to establish a sample and key feature library according to the classification data;

[0143] The performance analysis module 16 is used to calculate the evaluation index and statistical characteristics of the cycle according to the samples and the key feature library to achieve performance analysis.

[0144] According to the performance analysis system of the linear internal combustion power generation system disclosed in the above embodiment, the functions to be implemented by each module are implemented according to the steps in the performance analysis method of the linear internal combustion power generation system disclosed in the above embodiment, which will not be repeated here.

[0145] The present invention also provides a storage medium storing a computer program. When the computer program is executed, it can implement the performance analysis method of the linear internal combustion power generation system disclosed in any one of the above-mentioned embodiments.

[0146] Among them, the data generated by the computer program exists in the form of cell arrays and arrays, which can be directly connected to the relevant neural network or correlation analysis software for rapid subsequent processing.

[0147] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A performance analysis method for a linear internal combustion power generation system, characterized in that: include: Collecting operation data of the linear internal combustion power generation system and preprocessing the operation data; For the pre-processed running data, a reasonable position is selected to divide it into loop segment data, and the specific process includes: Determine whether the speed change range and position change range before and after the preset segmentation point are within the preset threshold range. If not, use a genetic algorithm to find a segmentation point that meets the preset conditions. If so, use the preset segmentation point. Dividing the running data into a plurality of cycles according to the characteristics of the segmentation points, and determining whether there are repeated segmentations and missed segmentations; If it is determined that there are repeated segmentations and missed segmentations, the operation data is re-circularly divided according to the characteristics of the segmentation points; Combining the loop segment data of the same loop to establish a sample library; The key operation features are extracted from the cyclic sample data in the sample library, and the specific process includes: The ignition timing position is determined according to the characteristics of the cyclic ignition signal, and the cylinder pressure and speed at the ignition timing are recorded; According to the cycle sample data, finding and recording the peak cylinder pressure and the corresponding position of the peak cylinder pressure; According to the cycle sample data, finding and recording the top dead center and the bottom dead center; According to the cycle sample data, finding and recording the expansion point speed and the expansion point cylinder pressure; According to the cycle sample data, calculate and record the intake signal jump time, intake temperature and intake pressure; The cycles are classified and judged according to the extracted key operation features, and samples and key feature libraries are established according to the classification data. The specific process includes: According to whether the difference between the cylinder pressure at the expansion point and the cylinder pressure at the ignition time is less than 0, it is determined whether the current cycle is a misfire cycle, and all data and feature arrays of the current cycle belonging to the misfire cycle are stored in the misfire cycle sample and key feature library; According to whether the difference between the top dead center and the stall preset value is less than 0, it is determined whether the current cycle is a stall cycle, and all data and feature arrays of the current cycle belonging to the stall cycle are stored in the stall cycle sample and key feature library; According to whether the difference between the expansion point speed and the preset value of the expansion point speed is less than 0, it is judged whether the current cycle is an incomplete combustion situation, and all data and feature arrays of the current cycle belonging to the incomplete combustion situation are stored in the incomplete combustion situation sample and key feature library, and, according to whether the difference between the ratio of the cylinder pressure and the intake pressure at the ignition time and the preset compression ratio is less than 0, it is further judged that the current incomplete combustion situation is a compression ratio insufficient cycle or a quenching cycle, and the compression ratio insufficient cycle and the quenching cycle are stored in the compression ratio insufficient cycle sample and key feature sub-library and the quenching cycle sample and key feature sub-library respectively; According to whether the difference between the cylinder pressure at the ignition time and the abnormal cycle monitoring value is less than 0, it is determined whether the current cycle is other abnormal cycles, and all data and feature arrays of the current cycle belonging to other abnormal cycles are stored in other abnormal cycle samples and key feature libraries; Store all data and feature arrays of the current cycle of normal cycles that do not belong to misfire cycles, stall cycles, incomplete combustion conditions, and other abnormal cycles into a normal cycle sample and key feature library; According to the samples and key feature library, the evaluation index and statistical characteristics of the cycle are calculated to achieve performance analysis. The specific process includes: For different types of cycles, the cycle frequency, cycle indicated work and cycle thermal efficiency are calculated as evaluation indicators; The calculation formula of cycle efficiency Fre is: Among them, Δt is the time used in the current cycle; The calculation formula of cycle indicated work is: ∫1 km P m Tox Among them, P m is the cylinder pressure of the mth cycle, A is the piston area in the cylinder, 1 and km are the starting and ending points of the data of the mth cycle; For different types of cycles, the average value of top dead center, the average value of bottom dead center, the average value of peak cylinder pressure, the average value of frequency and the average value of indicated work as well as the mean square error of top dead center and the mean square error of peak cylinder pressure are calculated as statistical features. The calculation formula for the average value μ is: Mean square error σ 2 The calculation formula is: Where n is the number of cycles involved in the average value calculation, c i is the eigenvalue involved in the calculation cycle.

2. The performance analysis method of a linear internal combustion power generation system according to claim 1, characterized in that: The specific process of preprocessing the operation data includes: Determine whether there is motion speed data in the operation data, and if not, obtain the motion speed data by differentiating the position data and the time data; Performing Gaussian filtering on the motion speed data to obtain relatively smooth motion speed data; The cylinder pressure data in the operation data is filtered.

3. The performance analysis method of a linear internal combustion power generation system according to claim 1, characterized in that: The loop fragment data belonging to the same loop are divided into multiple loop fragments, which are respectively entered into cell arrays for storage to establish a sample library.

4. The performance analysis method of a linear internal combustion power generation system according to claim 1, characterized in that: Also includes: Based on the calculated evaluation indicators and statistical characteristics of the cycle, a statistical graph characterizing the cycle performance is drawn.

5. A performance analysis system for a linear internal combustion power generation system, characterized in that: The performance analysis method of the linear internal combustion power generation system according to any one of claims 1 to 4 comprises: A data preprocessing module, used for collecting operation data of the linear internal combustion power generation system and preprocessing the operation data; A loop division module is used to select reasonable positions to divide the pre-processed running data into loop segment data; A sample building module is used to combine loop segment data of the same loop to build a sample library; A feature extraction module, used for extracting key operation features from the cyclic sample data in the sample library; A cycle classification module is used to classify and judge the cycle according to the extracted key operation characteristics, and to establish a sample and key feature library according to the classification data; The performance analysis module is used to calculate the evaluation index and statistical characteristics of the cycle according to the samples and key feature library to achieve performance analysis.

6. A storage medium storing a computer program, characterized in that: When the computer program is executed, the performance analysis method of the linear internal combustion power generation system according to any one of claims 1 to 4 can be implemented.

Images