Method for correcting combustion starting point identification deviation based on combustion excitation contribution degree
Technical Field
The invention belongs to the technical field of internal combustion engines, and particularly relates to a method for correcting combustion starting point identification deviation based on combustion excitation contribution degrees.
Background
The internal combustion engine vibration signal contains a large amount of information related to combustion, and is widely used for extracting a combustion excitation response signal. However, the vibration signal also contains non-combustion excitation sources such as piston reversing impact force, piston side pressure, reciprocating inertia force and the like, and the non-combustion excitation sources and the combustion excitation are mutually coupled in time domain and frequency domain, so that the difficulty in extracting combustion information is increased. And the high-frequency harmonic components in the in-cylinder pressure signals under different working conditions are different, so that the phase between the vibration signal and the excitation signal is changed, and the phase lag angle exists between the phase combustion characteristic parameter identified from the vibration signal and the characteristic point calculated based on the in-cylinder pressure signals.
In the prior art, the relation between a cylinder head surface vibration acceleration signal and the second derivative of the pressure in a cylinder is analyzed in a simulation mode in a text of identifying the combustion starting point of the HCCI engine by using the vibration acceleration. The result shows that before the peak pressure in the cylinder appears, the vibration acceleration signal of the surface of the cylinder cover and the second derivative of the pressure in the cylinder have approximate change rules. And aiming at the condition that the combustion starting point identified by the vibration acceleration signal lags behind the combustion starting point represented by the pressure second derivative, the lag angle is regarded as system lag, and the combustion starting point phase identification precision is improved. The lag angle is corrected through system deviation, the phase combustion characteristic parameter identification precision under partial working conditions can be improved, but the physical meaning is lacked, and the phase lag angles under different working conditions are different.
Chinese patent (CN108875581A) discloses a method for describing the phase lag angle between the vibration speed signal and the in-cylinder pressure signal by using the characteristic parameters in the surface vibration speed signal of the internal combustion engine. The method comprises the steps of firstly eliminating high-frequency interference signals and low-frequency interference signals in vibration speed signals by using a low-pass filter and a wavelet decomposition technology, then finding out characteristic parameters representing and describing phase lag from the reconstructed vibration speed signals, and substituting the parameters into corresponding formulas to describe lag angles between the vibration speed signals and excitation signals. However, the method describes the lag angle by using the characteristic parameter in the vibration speed signal to represent the characteristic parameter in the cylinder pressure signal, is easily influenced by the reconstruction error of the vibration speed signal, and has a complex whole description process.
Chinese patent (CN111964913A) discloses a contribution calculation method, which uses a partial coherence function to solve the ratio of the frequency spectrum amplitude of a vibration velocity signal when a main excitation is applied alone to the amplitude of the vibration velocity signal when all excitations are applied at a certain frequency, and uses the partial coherence function to calculate the contribution of the excitation signal to the vibration signal at the specific frequency. The contribution degree of the invention is that under a certain working condition, after the vibration signal of the cylinder cover surface is separated by a Principal Component Analysis (PCA) method, the ratio of the characteristic value of each principal Component to the sum of all the characteristic values is simpler and more convenient in calculation process, and the calculation adaptability of the contribution degree relative to a specific frequency is higher. Meanwhile, the calculation of the contribution degree indicated in the Chinese patent (CN111964913A) is the basis for researching and removing the interference in the vibration signal and is beneficial to repairing partial combustion information in the vibration signal; the contribution calculation of the invention aims to realize the correction of the combustion initial point phase deviation and is beneficial to improving the identification precision of the combustion characteristic parameters under all working conditions.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides a method for correcting combustion starting point identification deviation based on combustion excitation contribution, which is based on vibration signal characteristic analysis, utilizes a principal component analysis method to extract a combustion excitation response signal from a vibration signal and calculate the contribution degree of the combustion excitation response signal, establishes a phase deviation correction curve according to the change rule between the phase deviation of the combustion starting point and the contribution level under different working conditions, and further realizes the correction of the combustion starting point lag angle.
The present invention achieves the above-described object by the following technical means.
A method of correcting a combustion start point identification deviation based on a combustion excitation contribution degree, comprising the steps of:
s1, determining a main excitation source in the vibration acceleration signal;
s2, extracting a combustion excitation response signal from the vibration acceleration signal by using a principal component analysis method;
and S3, calculating the contribution level of the combustion excitation response signal under different working conditions and the corresponding combustion starting point phase deviation, establishing a phase deviation correction curve, and further realizing the correction of the combustion starting point phase deviation.
In the above technical solution, the determining the main excitation source in the vibration acceleration signal specifically includes:
establishing a finite element model of the two-cylinder diesel engine, simulating the vibration characteristics of different excitation sources under the independent action, and comparing the maximum amplitude of the vibration acceleration of each excitation source under the independent action.
In the technical scheme, the excitation source comprises the pressure in the cylinder, the piston reversing impact force, the piston side pressure and the reciprocating inertia force.
In the above technical solution, the S2 specifically is: selecting vibration acceleration signals recorded at a plurality of position points above a cylinder cover of the two-cylinder diesel engine to form a data set for principal component analysis, calculating a covariance matrix, a characteristic value and a characteristic vector of the standardized data set, and multiplying the original data set by the characteristic vector to obtain each principal component regression curve after separation; and determining a combustion excitation response signal according to the change rule that the principal component regression curve and the second derivative curve of the in-cylinder pressure have the similar change rule.
In the above technical solution, the contribution level of the combustion excitation response signal is: and the ratio of the characteristic value corresponding to the characteristic vector representing the combustion excitation response signal to the sum of all the characteristic values is the contribution level of the combustion excitation response signal.
In the above technical solution, the phase deviation of the combustion starting point is: the crank angle corresponding to the first zero-crossing point before the peak value of the combustion excitation response signal is APCAUnder the corresponding working condition, the crank angle corresponding to the first zero crossing point before the peak value of the second derivative curve of the in-cylinder pressure is AICPThe phase deviation of the combustion start point is Adeviation=APCA-AICP。
In the above technical solution, the phase deviation correction curve is: and according to the change rule between the contribution level of the combustion excitation response signal and the corresponding combustion initial point deviation under different working conditions, establishing a third-order polynomial fitting curve.
In the above technical solution, the correction of the combustion starting point phase deviation is: on-line measurement of vibration acceleration at different position pointsThe signals form a sample set, then principal component analysis processing is carried out, combustion excitation response signals are extracted, and corresponding combustion initial point phase deviation A is calculateddeviationAnd a contribution level CICPWill contribute to level CICPBringing the phase deviation correction curve into the established phase deviation correction curve to find the corresponding correction angle AcorrectThe corrected combustion start point phase deviation is Acorrected=Adeviation-Acorrect。
The invention has the beneficial effects that: the combustion excitation response signal is extracted from the vibration acceleration signal, the combustion starting point is extracted based on the combustion excitation response signal, the contribution level of the extracted combustion excitation response signal under different working conditions and the corresponding phase deviation of the combustion starting point are calculated, and a phase deviation correction curve is established based on the change rule between the contribution level of the combustion excitation response signal and the corresponding phase deviation of the combustion starting point, so that the correction of the phase deviation of the combustion starting point is realized, and the identification accuracy of the combustion characteristic parameters under the full working conditions is improved.
Drawings
FIG. 1(a) is a simulation model diagram of a two-cylinder diesel engine according to the present invention;
FIG. 1(b) is a schematic diagram of a vibration acceleration signal output position point of a two-cylinder diesel engine according to the present invention;
FIG. 2(a) is a graph of in-cylinder pressure vibration acceleration according to the present invention;
FIG. 2(b) is a graph of piston reversing impact force vibration acceleration in accordance with the present invention;
FIG. 2(c) is a graph of piston side pressure oscillation acceleration in accordance with the present invention;
FIG. 2(d) is a graph of the reciprocating inertial force vibration acceleration of the present invention;
FIG. 3 is a graph comparing the extracted combustion excitation response signal and the second derivative curve of in-cylinder pressure according to the present invention;
FIG. 4(a) is a statistical graph of the contribution at a torque of 40 N.m and at speeds of 800rpm, 1200rpm, 1650rpm, 2200rpm, respectively;
FIG. 4(b) is a statistical chart of the contribution degrees at a rotation speed of 1650rpm and torques of 0N · m, 25N · m, 50N · m and 100N · m, respectively;
FIG. 5 is a third order polynomial fit plot of combustion onset phase deviation as a function of contribution level in accordance with the present invention;
FIG. 6 is a comparison chart of the combustion start point phase deviation before and after correction according to the present invention;
FIG. 7 is a flow chart of a method of correcting combustion onset identification bias based on combustion excitation contribution of the present invention.
Detailed Description
The invention will be further described with reference to the following figures and specific examples, but the scope of the invention is not limited thereto.
As shown in fig. 7, a method for correcting a combustion start point identification deviation based on a combustion excitation contribution degree specifically includes the following steps:
the method comprises the following steps: determining a main excitation source in the vibration acceleration signal;
the method for determining the main excitation source comprises the following steps: establishing a finite element model of the two-cylinder diesel engine, simulating the vibration characteristics of different excitation sources under the independent action, and comparing the maximum amplitude of the vibration acceleration of each excitation source under the independent action;
as shown in figure 1(a), in order to determine the vibration characteristics generated under the independent action of each main excitation source, the invention establishes a two-cylinder diesel engine finite element model taking a machine body and a cylinder cover as main bodies in Abaqus, adopts 10-node tetrahedral meshes, and simultaneously establishes 4 rigid supports which do not deform, wherein each node on each support is restrained, so that the freedom degree of each direction of each support is limited. Fig. 1(b) shows the vibration acceleration signal output positions above the cylinder head, wherein a1-a7 indicate vibration acceleration signal output position points at different positions.
As shown in fig. 2(a), (b), (c) and (d), the vibration acceleration diagram generated above the cylinder head is generated under the action of four excitation sources of cylinder pressure, piston reversing impact force, piston side pressure and reciprocating inertia force when the rotating speed is 1650rpm and the torque is 40N · m. In order to reduce the influence of excitation of adjacent cylinders, a vibration acceleration signal in a range of 90 degrees CA before and after the combustion top dead center of one cylinder is intercepted, the maximum amplitude of the vibration acceleration of the piston impact force, the piston side pressure and the reciprocating inertia force in the interval is respectively 20.7 percent, 4.1 percent and 5.2 percent of the amplitude of the vibration acceleration of the pressure in the cylinder, the influence of the piston reversing impact force cannot be ignored, the energy of the reciprocating inertia force can be rapidly increased along with the increase of the rotating speed, further analysis is needed, and therefore in subsequent analysis, the four excitations are applied.
Step two: extracting a combustion excitation response signal from the vibration acceleration signal by using a principal component analysis method;
selecting vibration acceleration signals (the position points are more than the number of excitation sources) recorded at a plurality of position points (A1-A7) above a cylinder cover to form a data set of Principal Component Analysis (PCA), standardizing the data set, calculating a covariance matrix, a characteristic value and a characteristic vector of the standardized data set, and multiplying the original data set and the characteristic vector to obtain each separated principal Component regression curve.
As shown in fig. 3, the principal component regression curve has a similar variation law with the second derivative curve of the in-cylinder pressure, and it can be determined that the component is caused by the in-cylinder pressure excitation. In terms of phase, the principal component regression curve slightly lags behind the second derivative curve of the cylinder pressure, which is the phase change between the vibration acceleration signal and the excitation signal caused by different high-frequency harmonic components in the cylinder pressure signal under different working conditions, so that after the principal component analysis is performed by using the vibration acceleration signal, each principal component regression curve also has a certain lag.
Step three: and calculating the contribution level of the combustion excitation response signal extracted in the step two under different working conditions and the corresponding combustion initial point phase deviation, establishing a phase deviation correction curve, and further realizing the correction of the combustion initial point phase deviation.
The contribution level of the combustion excitation response signal refers to the weight of the combustion excitation response signal in the vibration acceleration signal; the combustion initial point phase deviation refers to the difference between a combustion excitation response signal curve extracted by a PCA method and a crank shaft rotation angle corresponding to a first zero-crossing point before the peak value of a corresponding in-cylinder pressure secondary derivative curve; the phase deviation correction curve refers to a third-order polynomial fitting curve between the contribution level of the combustion excitation response signal and the phase deviation of the corresponding combustion starting point under multiple groups of working conditions.
The PCA method can synchronously calculate the contribution degree of each principal component in the process of separating the vibration acceleration signals. As shown in FIG. 4, statistics of the contribution of the combustion excitation response signal are displayed. Fig. 4(a) shows the main component contribution degrees of the combustion excitation response signal under the operating conditions of 800rpm, 1200rpm, 1650rpm and 2200rpm at 40N · m, and fig. 4(b) shows the main component contribution degrees of the combustion excitation response signal under the operating conditions of 0N · m, 25N · m, 50N · m and 100N · m at 1650 rpm. As can be seen from the graph, the level of contribution of the combustion excitation response signal in the vibration acceleration signal gradually decreases with increasing rotational speed, and slightly increases with increasing torque. The reason is that the increase of the rotating speed can lead to the increase of the non-combustion excitation intensity applied to the finite element model of the two-cylinder diesel engine, thereby reducing the signal-to-noise ratio of the pressure signal in the cylinder; at the rated speed, the increase of the torque increases the fuel injection quantity, thereby improving the energy of the pressure in the cylinder. The lag angle of the phase characteristic parameter can also change along with the change of the working condition, and a phase deviation correction curve can be established according to the change rule between the lag angle of the phase characteristic parameter and the contribution level of the combustion excitation response signal.
The specific method for establishing the phase deviation correction curve is as follows:
firstly, calculating the phase deviation of a combustion starting point; a crank angle corresponding to the first zero-crossing point before the peak value of the combustion excitation response signal extracted in the step two is APCAThe crank angle corresponding to the first zero crossing point before the peak of the second derivative curve of the in-cylinder pressure under the corresponding working condition is AICPThe phase deviation of the combustion start point is Adeviation=APCA-AICP。
Secondly, calculating the contribution level of the combustion excitation response signal; measuring vibration acceleration signals of different position points from the surface of the cylinder cover, forming the vibration acceleration signals into an n-dimensional sample set X, and obtaining n characteristic vectors U representing the original sample set after the sample set X is processed by PCAiAnd corresponding characteristic value lambdai(i ═ 1, 2, …, n); a sample set X and a feature vector U are combinediMultiplying to obtain regression curve X of each main component after separationR=X·UiWave form and phase in each regression curve and cylinder internal pressureThe similar force second derivative curves are combustion excitation response signals; characteristic value lambda of regression curve corresponding to combustion excitation response signalICPWith the sum of all characteristic values λSUMI.e. the contribution level C of the combustion excitation response signal in the vibration acceleration signalICP=λICP/λSUM。
Finally, according to the change rule between the contribution level of the combustion excitation response signal and the corresponding combustion initial point deviation under different working conditions, a third-order polynomial fitting curve is established; four torque moments of 20 N.m, 40 N.m, 60 N.m and 80 N.m are selected under four rotating speeds of 800r/min, 1200r/min, 1650r/min and 2200r/min to form 16 groups of working conditions, and the contribution level of the combustion excitation response signal and the phase deviation of the combustion starting point calculated under the working conditions form 16 groups of coordinates as the reference points of a fitting curve, so that a third-order polynomial fitting curve is established, and the fitting curve can effectively reflect the change rule of the phase deviation along with the contribution level, see fig. 5.
The method for correcting the combustion starting point phase deviation comprises the following steps: measuring vibration acceleration signals of different position points on line, forming the vibration acceleration signals into a sample set, carrying out PCA (principal component analysis) processing, extracting combustion excitation response signals and calculating phase deviation A of combustion starting points of the combustion excitation response signalsdeviationAnd a contribution level CICPWill contribute to level CICPBringing the phase deviation correction curve (namely a third-order polynomial fitting curve) into the well established phase deviation correction curve, and finding out the corresponding correction angle AcorrectThe corrected combustion start point phase deviation is Acorrected=Adeviation-Acorrect。
As shown in fig. 6, a comparison chart before and after the combustion start point phase deviation correction; it can be seen from the figure that after the phase deviation of the combustion starting point under each working condition is corrected by a third-order polynomial fitting curve, the deviation value is remarkably reduced, and the maximum value of the corrected deviation angle is 0.44-degree CA, so that the method can effectively correct the phase deviation of the combustion starting point.
The present invention is not limited to the above-described embodiments, and any obvious improvements, substitutions or modifications can be made by those skilled in the art without departing from the spirit of the present invention.