A method and system for detecting deactivation camshaft displacement

Abstract

The application discloses a kind of for the detection method and system of deactivation camshaft displacement, belong to engine deactivation control technical field.The method includes: obtaining displacement original signal output by deactivation camshaft displacement sensor and the rotation angle synchronous signal output by crankshaft rotation angle sensor;The intrinsic mode function set is obtained by experience mode decomposition to displacement original signal, and angle domain resampling is carried out using rotation angle synchronous signal, to obtain angle domain mode function set;From angle domain mode function set, extract displacement mutation feature and convert into displacement state symbol sequence;According to displacement state symbol sequence, construct Markov transition field matrix, calculate spectral radius;Spectral radius is compared with preset spectral radius threshold interval, and camshaft displacement detection result is output.The scheme can reduce the influence of speed fluctuation and noise interference on displacement detection, improve the accuracy of deactivation camshaft displacement state discrimination.

Description

Technical Field

[0001] This invention relates to the field of engine cylinder deactivation control technology, specifically to a method and system for detecting camshaft displacement during cylinder deactivation. Background Technology

[0002] Cylinder deactivation camshaft displacement detection is used to determine whether the cylinder deactivation mechanism has completed the camshaft axial position switching according to control commands. Existing detection methods typically acquire the output signal of a displacement sensor and perform state discrimination based on time-domain amplitude thresholds, rate of change thresholds, or envelope features. Some schemes use frequency domain analysis methods such as Fourier transform and wavelet transform to extract features. Cylinder deactivation camshaft displacement signals often exhibit transient abrupt changes synchronized with crankshaft rotation angle. The displacement sensor output is significantly affected by engine speed fluctuations, vibration noise, and electromagnetic interference. Time-domain thresholding methods are prone to misjudgment when engine speed changes or background noise is strong. Frequency domain analysis methods have limited ability to locate non-stationary abrupt changes in time and cannot accurately describe the transition process between displacement states. If the crankshaft rotation angle synchronization signal and the displacement signal are only time-aligned or only used for trigger sampling without angle domain resampling, speed fluctuations will cause inconsistent sampling points for the same camshaft displacement event in different cycles, affecting the consistency of subsequent feature extraction. Existing methods lack symbolic and quantitative expressions for displacement abrupt changes and state transitions, usually obtaining local amplitude or frequency indices, which cannot effectively reflect the transition structure and degree of anomaly of the displacement state sequence. The aforementioned issues result in insufficient consistency and accuracy of cylinder deactivation camshaft displacement detection under transient conditions and complex noise environments. The consistency of displacement signal angular domain sampling under speed fluctuation conditions and the comparable quantitative discrimination of displacement abrupt state transitions still need improvement. Summary of the Invention

[0003] To address the issues of time-domain thresholding and frequency-domain analysis methods being affected by speed fluctuations and noise interference, as well as the lack of quantitative discrimination for sudden displacement state transitions in cylinder-stopped camshaft displacement detection, this paper proposes a method and system for detecting cylinder-stopped camshaft displacement. The method decomposes and resamples the displacement signal within the crankshaft rotation domain, constructs a Markov transition field matrix using a sequence of displacement state symbols, and outputs the camshaft displacement detection results using the spectral radius as a discrimination index, thereby improving the consistency and accuracy of displacement state identification under transient conditions.

[0004] To achieve the above objectives, the present invention provides the following technical solution: The present invention provides a method and system for detecting camshaft displacement during cylinder deactivation. The method simultaneously acquires the raw displacement signal output by the camshaft displacement sensor and the synchronous angle signal output by the crankshaft angle sensor; performs empirical mode decomposition on the raw displacement signal to obtain an intrinsic mode function set; resamples the intrinsic mode function set in the angular domain using the synchronous angle signal to obtain an angular domain mode function set; extracts displacement abrupt change features from the angular domain mode function set and converts these features into a displacement state symbol sequence; constructs a Markov transfer field matrix based on the displacement state symbol sequence and calculates the spectral radius of the Markov transfer field matrix; compares the spectral radius with a preset spectral radius threshold range and outputs the detection result of the camshaft displacement. This method suppresses the influence of speed fluctuations through angular domain resampling and utilizes the spectral radius of the Markov transfer field matrix to reflect the changing characteristics of the displacement state, effectively identifying camshaft displacement anomalies during cylinder deactivation and improving the accuracy and robustness of detection.

[0005] As a preferred embodiment of the present invention, when performing empirical mode decomposition on the original displacement signal, extreme point detection is performed on the original displacement signal, and the upper and lower envelope mean values ​​are obtained through cubic spline interpolation. Eigenmode functions of each order are iteratively selected to form the eigenmode function set. The rotation synchronization signal is converted into an equal-angle increment sequence, and each eigenmode function in the eigenmode function set is interpolated and resampled based on the equal-angle increment sequence to generate the angular domain mode function set. This process transforms the displacement signal from the time domain to the angular domain, which can suppress signal non-stationarity caused by rotational speed changes and make feature extraction more stable.

[0006] As another preferred technical solution of the present invention, when extracting displacement abrupt change features from the angular domain modal function set, the local energy density of each modal function in the angular domain modal function set is calculated, the angular domain interval where the local energy density exceeds the energy threshold is located, and the amplitude change rate within the angular domain interval is used as the displacement abrupt change feature; the displacement abrupt change feature is encoded according to the sign and numerical range of the amplitude change rate to generate the displacement state symbol sequence. Through energy localization and symbol encoding, the displacement abrupt change caused by cylinder shutdown switching can be highlighted, redundant information can be compressed, and the efficiency and consistency of state representation can be improved.

[0007] As a further preferred technical solution of the present invention, when constructing the Markov transition field matrix based on the displacement state symbol sequence, the displacement state symbol sequence is divided into multiple time windows, and the transition probabilities between different symbols within each time window are statistically analyzed to form the Markov transition field matrix. The Markov transition field matrix is ​​then subjected to eigenvalue decomposition to extract all eigenvalues, and the maximum value of the modulus of the eigenvalues ​​is taken as the spectral radius. The spectral radius reflects the transition intensity and stability of the state symbol sequence and is closely related to the changes in the camshaft displacement state, which is beneficial for improving the sensitivity and anti-interference ability of anomaly detection.

[0008] As another preferred embodiment of the present invention, when comparing the spectral radius with a preset spectral radius threshold range, if the spectral radius falls within the preset spectral radius threshold range, a normal displacement signal is generated; if the spectral radius exceeds the preset spectral radius threshold range, an abnormal displacement signal is generated. Based on the normal displacement signal or the abnormal displacement signal, and combined with the maximum continuous symbol length in the displacement state symbol sequence, the detection result of the camshaft displacement is output. Combining the spectral radius and the maximum continuous symbol length for comprehensive judgment can further distinguish states such as displacement deviation, vibration, and incomplete cylinder stop switching, improving the interpretability of the detection results.

[0009] As a specific technical solution of the present invention, when acquiring the original displacement signal and the rotation angle synchronization signal, the voltage signal output by the cylinder deactivation camshaft displacement sensor and the pulse signal output by the crankshaft rotation angle sensor are synchronously acquired at a preset sampling frequency; the voltage signal is amplified and filtered to obtain the original displacement signal; the pulse signal is shaped and frequency multiplied to obtain the rotation angle synchronization signal. This method can improve the signal-to-noise ratio and angular resolution, providing high-quality data for subsequent decomposition and resampling.

[0010] As another specific technical solution of the present invention, when interpolating and resampling each intrinsic mode function in the intrinsic mode function set based on the equal-angle increment sequence, the starting point of the equal-angle increment sequence is aligned with the starting point of the intrinsic mode function set to determine the angular domain sampling time corresponding to each equal-angle increment; at the angular domain sampling time, a piecewise cubic conformal interpolation algorithm is used to interpolate the amplitude of each intrinsic mode function to obtain the angular domain mode function set. Piecewise cubic conformal interpolation can avoid oscillations and overshoot, ensuring the accuracy and stability of angular domain resampling.

[0011] As another specific technical solution of the present invention, when calculating the local energy density, the angular domain modal function set is divided into angular domain windows, the sum of squares of the amplitudes of all modal functions within each angular domain window is calculated, and the sum of squares is divided by the length of the angular domain window to obtain the local energy density; the local energy density is compared with the energy threshold, and continuous angular domain windows exceeding the energy threshold are retained as the angular domain interval; the ratio of the amplitude difference at the two endpoints of the angular domain interval to the angular domain span is calculated as the displacement mutation feature. Through local energy localization and amplitude change rate calculation, the location and intensity of the displacement mutation of the cylinder-stopping camshaft can be accurately extracted.

[0012] As another preferred technical solution of the present invention, when forming the Markov transition field matrix by statistically analyzing the transition probabilities between different symbols within each time window, each symbol in the displacement state symbol sequence is mapped to an index of the Markov transition field matrix; for each time window, the frequency of adjacent symbol pairs is counted, and the frequency is divided by the total frequency of all adjacent symbol pairs within that time window to obtain the transition probability; the corresponding positions of the Markov transition field matrix are filled according to the order of the indices to complete the construction of the Markov transition field matrix. This construction method can adaptively reflect the transition rules of the symbol sequence, reduce the dependence on manual thresholds, and improve the adaptability and stability of the overall detection method.

[0013] This invention also provides a detection system for cylinder deactivation camshaft displacement, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of the cylinder deactivation camshaft displacement detection method provided by this invention. This system can acquire, process, extract features, and determine the state of camshaft displacement signals in engine cylinder deactivation control scenarios, and features accurate detection and timely response.

[0014] The technical effects and advantages provided by the present invention in the above technical solution are as follows: Empirical Mode Decomposition (EMD) is performed on the original displacement signal to obtain a set of intrinsic mode functions (EMFs) containing components of different scales. The EMF set is then resampled in the angular domain using the crankshaft angle synchronization signal, forming a set of angular domain EMFs with equal angular intervals. After the crankshaft angle synchronization signal is converted into an equal-angle increment sequence, the EMFs of each order are resampled at the same crankshaft angle position, and the number of angular domain sampling points corresponding to the same camshaft displacement event tends to be consistent across different cycles. This processing transforms the displacement signal from the time domain to the crankshaft angle domain, eliminating the influence of engine speed fluctuations on the signal sampling interval, and ensuring that the same camshaft displacement event has a consistent angular domain sampling expression under different speed conditions. Empirical Mode Decomposition separates abrupt components, gradual trends, and noise components in the displacement signal. The angular domain resampled mode functions retain the angular domain local features corresponding to camshaft displacement changes, reducing the interference of vibration noise and electromagnetic interference on the feature extraction process. Compared to directly thresholding or spectral analysis of time-domain displacement signals, angular domain mode function sets can more accurately locate the crankshaft angle where displacement abrupt changes occur, enhancing the repeatability of displacement features across cycles. Displacement abrupt change features are extracted from the angular domain mode function set and encoded into a displacement state symbol sequence. A Markov transition field matrix is ​​constructed based on the displacement state symbol sequence, and its spectral radius is calculated. Displacement abrupt change features are obtained through the local energy density and amplitude change rate of the angular domain mode functions. The local energy density locates the interval where the abrupt change occurs, while the amplitude change rate describes the direction and strength of the abrupt change. The symbol sequence compresses minute fluctuations in the continuous signal. The displacement state symbol sequence describes the transition process of the displacement state within the angular domain. The Markov transition field matrix represents the state transition probability in the symbol sequence in a matrix form, and the spectral radius reflects the overall strength of the state transition structure. This method transforms the detection of cylinder-stopped camshaft displacement from a single local amplitude comparison to a quantitative discrimination at the state transition level. It can effectively identify abnormal changes in displacement state, reduce misjudgments caused by local noise spikes or single-point threshold exceedances, and enhance the detection capability of abnormal displacement states of cylinder-stopped camshafts. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0016] Figure 1 This is a flowchart of the cylinder deactivation camshaft displacement detection method; Figure 2 This is a flowchart of the Markov transition field matrix construction and spectral radius extraction for the displacement state symbol sequence; Figure 3This is a schematic diagram of the local energy density and displacement abrupt change characteristics of the angular domain modal function of the camshaft with cylinder deactivation; Figure 4 This is a schematic diagram of the radius sample distribution and threshold of the displacement state spectrum of the cylinder-stopped camshaft. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] See Figure 1 This invention provides a method for detecting the displacement of a cylinder-deactivated camshaft. The method acquires the raw displacement signal output by a cylinder-deactivated camshaft displacement sensor and simultaneously acquires the angular synchronization signal output by a crankshaft angle sensor. Empirical mode decomposition (EMD) is performed on the raw displacement signal to obtain a set of intrinsic mode functions (EMFs). The EMFs are then resampled in the angular domain using the angular synchronization signal to obtain a set of angular domain mode functions (EMFs). Displacement abrupt change features are extracted from the angular domain EMFs and converted into a displacement state symbol sequence. A Markov transition field matrix is ​​constructed based on the displacement state symbol sequence, and the spectral radius of the Markov transition field matrix is ​​calculated. The spectral radius is compared with a preset spectral radius threshold range to output the detection result of the camshaft displacement.

[0019] In practice, extreme point detection is performed on the raw displacement signal. The raw displacement signal is obtained by amplifying and filtering the voltage signal output by the cylinder deactivation camshaft displacement sensor. Extreme point detection includes identifying local maxima and local minima in the raw displacement signal. A local maximum is defined as a point in the raw displacement signal whose amplitude is greater than the amplitudes of two adjacent points, and a local minimum is defined as a point in the raw displacement signal whose amplitude is less than the amplitudes of two adjacent points.

[0020] In some embodiments, the upper and lower envelope means are obtained by cubic spline interpolation. Cubic spline interpolation is performed on the detected local maxima to construct the upper envelope. Cubic spline interpolation is performed on the detected local minima to construct the lower envelope. The average values ​​of the upper and lower envelopes at each sampling point are calculated to obtain the upper and lower envelope mean curves.

[0021] The iterative process of selecting intrinsic mode functions (EMFs) of various orders can be understood as follows: Subtract the upper and lower envelope mean curves from the original displacement signal to obtain candidate EMFs. Determine whether the candidate EMFs satisfy the EMF conditions. The EMF conditions include: the number of extrema and zero-crossings are equal or differ by one over the entire signal length; and the average of the upper and lower envelopes is zero at any given time. If a candidate EMF does not satisfy the EMF conditions, it is used as a new input signal, and the processes of extrema detection, cubic spline interpolation to construct upper and lower envelopes, calculation of the upper and lower envelope mean curves, and subtraction of the upper and lower envelope mean curves are repeated until the candidate EMFs satisfy the EMF conditions. Candidate EMFs that satisfy the EMF conditions are designated as first-order EMFs. Subtract the first-order EMFs from the original displacement signal to obtain the residual signal. Using the residual signal as a new input signal, the process of extreme point detection, cubic spline interpolation to construct upper and lower envelopes, calculation of the upper and lower envelope mean, subtraction of the upper and lower envelope mean curves, and determination of eigenmode function conditions is repeated. Each order of eigenmode function is sequentially selected until the residual signal becomes a monotonic function or its amplitude is less than a preset residual threshold. The selected eigenmode functions are then combined into an eigenmode function set.

[0022] The crankshaft angle synchronization signal is converted into a constant angle increment sequence. The crankshaft angle synchronization signal is obtained by shaping and frequency multiplication of the pulse signal output from the crankshaft angle sensor. The crankshaft angle synchronization signal contains a pulse sequence corresponding to the crankshaft angle. The constant angle increment sequence is determined based on the pulse sequence number and pulse interval of the crankshaft angle synchronization signal. The constant angle increment sequence contains a set of equally spaced angle values, with a constant angle difference between adjacent angle values. The constant angle increment sequence is set to contain... Each corner domain sampling point, A positive integer. Obtain the starting angle of the crankshaft angle synchronization signal. and ending angle Starting angle The starting angle and ending angle of the equal-angle increment sequence. This represents the ending angle of the equal-angle increment sequence. Calculate the increment interval. : in, This is the ending angle of the crankshaft rotation synchronization signal. This is the starting angle of the crankshaft rotation angle synchronization signal. This represents the number of angular domain sampling points in the equal-angle increment sequence. The increment interval of the equal-angle increment sequence. The ending angle of the crankshaft angle synchronization signal. and starting angle The crankshaft angle is determined by the pulse sequence number and pulse resolution output by the crankshaft angle sensor. The crankshaft angle sensor pulse resolution corresponds to the crankshaft angle per pulse. The number of angle domain sampling points... The number of time sampling points for each intrinsic mode function in the intrinsic mode function set and the allowable interpolation error for speed fluctuations are set. The number of angular domain sampling points is also set. The larger the value, the higher the angular resolution of the equal-angle increment sequence, and the smaller the interpolation error introduced by the piecewise cubic conformal interpolation algorithm.

[0023] Based on the equal-angle increment sequence, each intrinsic mode function (EMF) in the EEMF set is interpolated and resampled to generate a angular domain EEMF set. The starting point of the equal-angle increment sequence is aligned with the starting point of the EEMF set. The starting point of the equal-angle increment sequence corresponds to the starting time sampling point of the EEMF set. The starting point of the EEMF set is the first sampling point of each EEMF in the set on the time series. After alignment, the angular domain sampling time corresponding to each equal-angle increment is determined. The angular domain sampling time is the time when the crankshaft angle position corresponding to each angle value in the equal-angle increment sequence is mapped onto the time axis. The angular domain sampling time is determined by establishing a correspondence between crankshaft angle and time based on the timestamp corresponding to each pulse in the crankshaft angle synchronization signal. For each angle value in the equal-angle increment sequence, the corresponding angular domain sampling time is obtained by interpolation through the correspondence between crankshaft angle and time. Optionally, the interpolation method can be linear interpolation or cubic spline interpolation.

[0024] At the angular domain sampling time, a piecewise cubic conformal interpolation algorithm is used to interpolate the amplitude of each intrinsic mode function (EMF), resulting in an angular domain EMF set. The piecewise cubic conformal interpolation algorithm uses a cubic polynomial for interpolation within each segment interval, and the interpolation curve remains monotonic at known data points to avoid overshoot. For each EMF in the EMF set, the amplitude sequence of the EMF at the time sampling points is obtained. Based on the correspondence between the angular domain sampling time and the time sampling points, the interpolation interval for each angular domain sampling time is determined. Within the interpolation interval, the piecewise cubic conformal interpolation algorithm is used to calculate the EMF amplitude at the angular domain sampling time. The EMF amplitudes at all angular domain sampling times are combined to form the angular domain EMF. The angular domain EMFs corresponding to each EMF in the EMF set are then combined to form the angular domain EMF set.

[0025] In practical implementation, the local energy density of each mode function in the angular domain modal function set is calculated, and the angular domain intervals where the local energy density exceeds the energy threshold are located. The amplitude change rate within the angular domain interval is used as the displacement abrupt change feature. The displacement abrupt change feature is encoded according to the sign and numerical range of the amplitude change rate to generate a displacement state symbol sequence.

[0026] The angular domain modal function set is divided into angular domain windows. Each angular domain modal function set contains multiple modal functions, and each modal function has an amplitude sequence with equal angular intervals over the angular domain. The angular domain window length is set. Corner window length This represents the number of corner sampling points contained within the corner window. Corner window length. Based on the total number of angular domain sampling points in the angular domain modal function set Sure, Pick The integer part, and the corner window length The minimum is 5 corner domain sampling points. The total number of corner domain sampling points in the corner domain modal function set. This represents the total number of angular domain sampling points for any mode function in the angular domain mode function set. Adjacent angular domain windows do not overlap, and the angular domain windows are arranged sequentially starting from the initial angular domain position of the angular domain mode function set, covering the entire angular domain mode function set.

[0027] Calculate the sum of squares of the magnitudes of all modal functions within each corner window, and divide the sum by the length of the corner window to obtain the local energy density. The formula for calculating the local energy density is: in, Local energy density; The length of the corner domain window is the number of corner domain sampling points contained within the corner domain window. This represents the total number of mode functions in the angular domain mode function set; This represents the number of corner sampling points within a single corner window. Equal to the length of the corner domain window ; For the first The modal function within the angular domain window is the first... The amplitude at each sampling point in the angular domain. The total number of modal functions in the angular domain modal function set. The number of eigenmode functions is determined by the total number of eigenmode functions in the eigenmode function set. The number of angular domain sampling points within the angular domain window. With the length of the corner domain window Consistent.

[0028] Calculate the arithmetic mean of the local energy density for all corner windows. Set energy threshold Energy threshold Take the arithmetic mean of the local energy density of all corner windows. The energy threshold coefficient is set to 2, which is twice that of the displacement abrupt change region. The basis for setting the energy threshold coefficient to 2 is that the local energy density in the displacement abrupt change region exceeds the local energy density in the stable region, and the angular domain window with a local energy density exceeding twice the arithmetic mean contains displacement abrupt change features.

[0029] The local energy density of each corner window is compared with the energy threshold. In comparison, the local energy density is preserved above the energy threshold. A continuous angular domain window is defined as the angular domain interval. A continuous angular domain window is defined as a angular domain where adjacent elements have local energy densities exceeding an energy threshold. The set of corner domain windows. If the local energy density of a corner domain window exceeds an energy threshold. Furthermore, the local energy density of adjacent corner windows also exceeds the energy threshold. If so, then adjacent corner domain windows will be merged into the same corner domain interval.

[0030] The ratio of the amplitude difference at the two endpoints of the angular domain interval to the angular domain span is calculated as a characteristic of the displacement abrupt change. The amplitude difference at the two endpoints of the angular domain interval is obtained by subtracting the sum of the amplitudes of all modal functions at the starting angular domain sampling point from the sum of the amplitudes of all modal functions at the ending angular domain sampling point of the interval. The angular domain span is the angle difference between the angular domain position corresponding to the ending angular domain sampling point and the angular domain position corresponding to the starting angular domain sampling point of the interval. The angular domain position is determined by the angle value of the angular domain sampling point in the equal-angle increment sequence.

[0031] Displacement abrupt change characteristics are encoded according to the sign and numerical range of the rate of change of amplitude. The sign of the rate of change of amplitude includes positive, negative, and zero. A first threshold for the rate of change of amplitude is set. Second amplitude change rate threshold First amplitude change rate threshold Less than the second amplitude change rate threshold First amplitude change rate threshold Take the arithmetic mean of the absolute values ​​of the rate of change of amplitude of all displacement abrupt change characteristics, and then use the second rate of change threshold. The arithmetic mean of the absolute values ​​of the amplitude change rates of all displacement abrupt change characteristics is summed with twice the standard deviation. When the absolute value of the amplitude change rate is less than the first amplitude change rate threshold... When the displacement abrupt change characteristic is encoded as a zero change symbol, the absolute value of the amplitude change rate is greater than or equal to the first amplitude change rate threshold. And less than the second amplitude change rate threshold When the amplitude change rate is positive, the displacement abrupt change feature is encoded as a positive low amplitude change sign; when the amplitude change rate is negative, the displacement abrupt change feature is encoded as a negative low amplitude change sign. When the absolute value of the amplitude change rate is greater than or equal to the second amplitude change rate threshold... If the rate of change of amplitude is positive, the displacement abrupt change feature is encoded as a positive high-amplitude change symbol; if the rate of change of amplitude is negative, the displacement abrupt change feature is encoded as a negative high-amplitude change symbol. All symbols corresponding to displacement abrupt changes are arranged in angular domain order to generate a displacement state symbol sequence.

[0032] See Figure 3 In the figure, the horizontal axis represents the angular domain location, ranging from 0° to 720°, and the vertical axis represents the local energy density, in volts squared (V). 2 The blue solid line represents the curve of local energy density as a function of angular domain position, and the red dashed line represents the energy threshold, which is 1.0V. 2 The figure is marked with multiple green upper triangles and purple lower triangles, representing positive and negative displacement abrupt changes, respectively. The orange-filled areas represent continuous angular domains where the local energy density exceeds the energy threshold.

[0033] The curve shows a peak in local energy density between approximately 130° and 180°, slightly exceeding the threshold, but without forming a continuous range exceeding the threshold; therefore, no angular region is formed at this point. Subsequently, between approximately 360° and 430°, the local energy density significantly exceeds the energy threshold, forming a distinct continuous over-limit range, highlighted by orange shading. Within this angular region, alternating positive and negative displacement abrupt changes indicate a significant rate of change in amplitude, with diverse signs of these changes. The highest local energy density peak reaches approximately 1.45V. 2 The values ​​are far above the threshold, indicating a significant shift abruptness within this angular domain. Apart from the two main peak regions mentioned above, the local energy density remains low in other angular domain locations, not exceeding the energy threshold and not forming angular domain regions.

[0034] In specific implementation, please refer to Figure 2 The displacement state symbol sequence is divided into multiple time windows. The displacement state symbol sequence contains five types of symbols: zero change symbol, positive low-amplitude change symbol, negative low-amplitude change symbol, positive high-amplitude change symbol, and negative high-amplitude change symbol. The length of the time window is set. Time window length Total number of symbols in the displacement state symbol sequence Divide by the number of time windows The integer quotient is determined. Number of time windows. Set to twice the number of symbol types, i.e. Set the number of time windows. The value of 10 is based on the fact that each of the five symbols corresponds to at least two time windows, used to statistically analyze the transitions between that symbol and other symbols within different time windows, thus ensuring that the transition probability statistics cover different stages of the displacement state symbol sequence. Each time window contains... The last time window contains fewer symbols than the length of the time window. There is no overlap between adjacent time windows. If , then it includes all remaining symbols.

[0035] In some embodiments, each symbol in the displacement state symbol sequence is mapped to an index of the Markov transition field matrix. Zero-change symbols are mapped to index 1, positive low-amplitude change symbols to index 2, negative low-amplitude change symbols to index 3, positive high-amplitude change symbols to index 4, and negative high-amplitude change symbols to index 5. The row and column indices of the Markov transition field matrix are determined according to the order of indices 1 to 5.

[0036] Optionally, for each time window, the frequency of adjacent symbol pairs is counted. An adjacent symbol pair is a pair of symbols that are adjacent in position within the time window. The total frequency of adjacent symbol pairs is equal to the number of symbols within the time window minus 1. Indicates index Symbols to index The frequency of adjacent symbol pairs of the symbol, where For the row index of the Markov transition field matrix, For the column index of the Markov transition field matrix, and All are integers ranging from 1 to 5. Frequency count Divide by the total frequency of all adjacent symbol pairs within that time window To obtain the transition probability Transition probability The calculation formula is: in, For index Symbols to index The transition probability of the sign, For index Symbols to index The frequency of adjacent symbol pairs of a symbol. This represents the total frequency of all adjacent symbol pairs within the time window. is the row index of the Markov transition field matrix, with values ​​ranging from 1 to 5. is the column index of the Markov transition field matrix, with values ​​ranging from 1 to 5.

[0037] It is understandable that the Markov transition matrix is ​​constructed by filling the corresponding positions in the indexed order. The Markov transition matrix is ​​a 5x5 square matrix. Line 1 The column elements are the transition probabilities. Transition probability The value range is from 0 to 1, and the sum of the values ​​of all elements in each row is equal to 1.

[0038] Eigenvalue decomposition is performed on the Markov transition field matrix to extract all eigenvalues. The QR algorithm is used for eigenvalue decomposition, which iteratively decomposes the Markov transition field matrix into the product of an orthogonal matrix and an upper triangular matrix. After convergence, the eigenvalues ​​of the Markov transition field matrix are obtained. Eigenvalues ​​are either real or complex numbers. The Markov transition field matrix is ​​a 5x5 square matrix with 5 eigenvalues. The modulus of each eigenvalue is calculated as follows: if the eigenvalue is real, the modulus is its absolute value; if the eigenvalue is complex, the modulus is the square root of the sum of the squares of its real and imaginary parts. The maximum modulus among all eigenvalues ​​is taken as the spectral radius.

[0039] In practical implementation, the spectral radius is obtained. The spectral radius is determined by the maximum modulus of all eigenvalues ​​of the Markov transition field matrix. A preset spectral radius threshold range is obtained. This preset threshold range includes a lower and an upper limit. The lower threshold is the minimum allowable value of the spectral radius under normal operating conditions, and the upper threshold is the maximum allowable value. The lower and upper thresholds are calculated by collecting multiple sets of spectral radius samples of the camshaft under normal displacement conditions with the cylinder deactivated, and then calculating the arithmetic mean of these multiple sets of spectral radius samples. and the standard deviation of multiple spectral radius samples Set the lower limit of the spectral radius threshold to Set the upper limit of the spectral radius threshold to The basis for setting the coefficient of 3 for both the lower and upper limits of the spectral radius threshold is that, under normal displacement conditions, the fluctuation range of the spectral radius covers the arithmetic mean plus or minus three standard deviations. Spectral radii exceeding this range are considered low-probability events and are judged as displacement anomalies. The preset spectral radius threshold interval is a closed interval, including both the lower and upper limits of the spectral radius threshold.

[0040] In some embodiments, the spectral radius is compared with a preset spectral radius threshold range. If the spectral radius is greater than or equal to the lower limit of the spectral radius threshold and less than or equal to the upper limit of the spectral radius threshold, the spectral radius is determined to fall within the preset spectral radius threshold range. If the spectral radius is less than the lower limit of the spectral radius threshold or greater than the upper limit of the spectral radius threshold, the spectral radius is determined to exceed the preset spectral radius threshold range.

[0041] Understandably, if the spectral radius falls within the preset spectral radius threshold range, a normal displacement signal is generated. This normal displacement signal is either a level signal or a digital flag, indicating that the displacement of the cylinder-stopped camshaft is within the normal fluctuation range. If the spectral radius exceeds the preset spectral radius threshold range, an abnormal displacement signal is generated. This abnormal displacement signal is either a level signal or a digital flag, indicating that the displacement of the cylinder-stopped camshaft deviates from the normal fluctuation range.

[0042] Based on the normal or abnormal displacement signal, and combined with the maximum continuous symbol length in the displacement state symbol sequence, the detection result of the camshaft displacement is output.

[0043] Obtain the maximum consecutive symbol length in the displacement state symbol sequence. The displacement state symbol sequence consists of zero-change symbols, positive low-amplitude change symbols, negative low-amplitude change symbols, positive high-amplitude change symbols, and negative high-amplitude change symbols arranged in angular domain order. The maximum consecutive symbol length is the maximum number of consecutive occurrences of the same symbol in the displacement state symbol sequence. The process of calculating the maximum consecutive symbol length is as follows: Set the current consecutive symbol length variable `current_run_length` and the maximum consecutive symbol length variable `max_run_length`, both with an initial value of 1. Starting from the second symbol in the displacement state symbol sequence, compare the current symbol with the previous symbol sequentially. If the current symbol is the same as the previous symbol, the current consecutive symbol length variable `current_run_length` is incremented by 1. If the current symbol is different from the previous symbol, the current consecutive symbol length variable `current_run_length` is compared with the maximum consecutive symbol length variable `max_run_length`. If the current consecutive symbol length variable `current_run_length` is greater than the maximum consecutive symbol length variable `max_run_length`, the maximum consecutive symbol length variable `max_run_length` is updated to the current consecutive symbol length variable `current_run_length`, and the current consecutive symbol length variable `current_run_length` is reset to 1. After traversing the displacement state symbol sequence, the current consecutive symbol length variable `current_run_length` is compared with the maximum consecutive symbol length variable `max_run_length`. If the current consecutive symbol length variable `current_run_length` is greater than the maximum consecutive symbol length variable `max_run_length`, the maximum consecutive symbol length variable `max_run_length` is updated to the current consecutive symbol length variable `current_run_length`. The final maximum consecutive symbol length variable `max_run_length` is the maximum consecutive symbol length.

[0044] A consecutive symbol length threshold is set. This threshold distinguishes the duration of consecutive identical symbols in a displacement state symbol sequence. The consecutive symbol length threshold is set to the integer part of one-tenth of the total number of symbols in the displacement state symbol sequence, with a minimum threshold of 10. The rationale for setting the consecutive symbol length threshold to one-tenth of the total number of symbols in the displacement state symbol sequence is as follows: under normal displacement conditions, displacement mutation characteristic symbols switch frequently, and the length of consecutive identical symbols does not exceed one-tenth of the total number of symbols; under abnormal displacement conditions, displacement mutation characteristic symbols remain unchanged, and the length of consecutive identical symbols exceeds one-tenth of the total number of symbols.

[0045] Output the detection results of camshaft displacement. If a normal displacement signal is generated and the maximum consecutive symbol length is less than or equal to the consecutive symbol length threshold, output the first detection result, indicating that the camshaft displacement is normal. If a normal displacement signal is generated and the maximum consecutive symbol length is greater than the consecutive symbol length threshold, output the second detection result, indicating that the camshaft displacement is normal but exhibits a trend of consecutive identical symbols. If an abnormal displacement signal is generated and the maximum consecutive symbol length is less than or equal to the consecutive symbol length threshold, output the third detection result, indicating that the camshaft displacement is abnormal and the abnormal features are scattered. If an abnormal displacement signal is generated and the maximum consecutive symbol length is greater than the consecutive symbol length threshold, output the fourth detection result, indicating that the camshaft displacement is abnormal and the abnormal features are continuous.

[0046] See Figure 4 In the graph, the horizontal axis represents the spectral radius value, and the vertical axis represents the frequency. The blue diagonal bar chart represents the distribution of spectral radius samples collected under normal displacement conditions, while the red diagonal bar chart represents the distribution of spectral radius samples collected under abnormal displacement conditions. The green dashed line represents the lower limit of the spectral radius threshold, and the purple dashed line represents the upper limit of the spectral radius threshold. The lower and upper limits of the spectral radius threshold are calculated from multiple sets of spectral radius samples under normal displacement conditions. Their values ​​are the arithmetic mean minus three standard deviations and the arithmetic mean plus three standard deviations, respectively, forming a closed interval range for the spectral radius threshold.

[0047] As shown in the figure, the spectral radius samples under normal displacement conditions are mainly concentrated between approximately 0.75 and 0.90, with a relatively high peak frequency. The vast majority of samples fall within the spectral radius threshold range, indicating that the spectral radius fluctuation is stable and limited in range under normal conditions. The spectral radius samples under abnormal displacement conditions have a wider distribution range, with many samples falling to the left of the lower limit of the spectral radius threshold, even as low as around 0.30. The frequency of samples outside the threshold range increases significantly, showing that the spectral radius deviates from the normal range under abnormal conditions. Furthermore, there is some overlap between normal and abnormal samples in the spectral radius value range, but the overall distribution of abnormal samples is left-skewed, with a larger number of samples falling below the lower threshold.

[0048] In practical implementation, the voltage signal output by the cylinder deactivation camshaft displacement sensor and the pulse signal output by the crankshaft angle sensor are synchronously acquired at a preset sampling frequency. The preset sampling frequency is determined based on the highest frequency component of the voltage signal output by the cylinder deactivation camshaft displacement sensor. The highest frequency component of the voltage signal output by the cylinder deactivation camshaft displacement sensor is the frequency corresponding to the rapid voltage fluctuation caused by the sudden displacement change during the cylinder deactivation camshaft displacement process. The preset sampling frequency is set to 8 times the highest frequency component of the voltage signal output by the cylinder deactivation camshaft displacement sensor. The basis for setting the preset sampling frequency to 8 times the highest frequency component is that the Nyquist sampling theorem requires the sampling frequency to be greater than twice the highest frequency component of the signal. To avoid spectral aliasing and retain the transient change characteristics in the original displacement signal, 8 times the highest frequency component is selected as the preset sampling frequency. Synchronous acquisition is achieved by simultaneously triggering the analog-to-digital conversion of the voltage signal output by the cylinder deactivation camshaft displacement sensor and the edge capture of the pulse signal output by the crankshaft angle sensor using a sampling clock. The sampling clock is generated by the timer of the detection system. The sampling clock generated by the timer is connected to the start pin of the analog-to-digital converter and the start pin of the pulse signal capture unit to ensure that the voltage signal sampling point and the pulse signal capture time are aligned on the same time base.

[0049] The voltage signal output from the cylinder deactivation camshaft displacement sensor is amplified and filtered to obtain the raw displacement signal. The voltage signal from the cylinder deactivation camshaft displacement sensor first enters an instrumentation amplifier for differential amplification. The gain of the instrumentation amplifier is determined based on the full-scale range of the output voltage signal from the cylinder deactivation camshaft displacement sensor and the input range of the analog-to-digital converter (ADC). The gain of the instrumentation amplifier is set to 0.9 times the ratio of the ADC input range to the full-scale range of the cylinder deactivation camshaft displacement sensor output voltage signal. The basis for setting the gain of the instrumentation amplifier to 0.9 times the ratio of the ADC input range to the full-scale range of the cylinder deactivation camshaft displacement sensor output voltage signal is to retain a 10% input margin and prevent the output voltage signal from exceeding the ADC input range under abnormal displacement conditions. The amplified voltage signal then enters a low-pass filter for filtering. The low-pass filter is a second-order Butterworth analog low-pass filter, and the cutoff frequency of the second-order Butterworth analog low-pass filter is set to 0.2 times the preset sampling frequency. The rationale for setting the cutoff frequency of the second-order Butterworth analog low-pass filter to 0.2 times the preset sampling frequency is as follows: the second-order Butterworth analog low-pass filter attenuates to -3dB at the cutoff frequency, effectively suppressing signals above the cutoff frequency and keeping signals below the cutoff frequency flat. Selecting 0.2 times the preset sampling frequency as the cutoff frequency meets the anti-aliasing filtering requirements under oversampling conditions. The voltage signal output by the low-pass filter is used as the original displacement signal.

[0050] The pulse signal output from the crankshaft angle sensor is shaped and frequency multiplied to obtain the angle synchronization signal. The pulse signal first enters a Schmitt trigger for shaping. The high and low threshold voltages of the Schmitt trigger are set based on the high and low level voltages of the pulse signal output from the crankshaft angle sensor. The high threshold voltage is set to 0.6 times the high level voltage of the crankshaft angle sensor's pulse signal, and the low threshold voltage is set to 0.4 times the low level voltage. The reason for setting the high threshold voltage to 0.6 times the high level voltage and the low threshold voltage to 0.4 times the low level voltage of the crankshaft angle sensor's pulse signal is that the hysteresis voltage between the high and low threshold voltages suppresses multiple triggers caused by noise in the pulse signal output from the crankshaft angle sensor. The shaped pulse signal then enters a frequency multiplier circuit for frequency multiplication. The frequency multiplier circuit uses a phase-locked loop (PLL) frequency multiplier. The phase-locked loop (PLL) frequency multiplier circuit includes a phase detector, a loop filter, a voltage-controlled oscillator (VCO), and a frequency divider. The phase detector compares the phase of the shaped pulse signal with the phase of the frequency divider's output signal. The phase error signal output by the phase detector is smoothed by the loop filter and then used to control the output frequency of the VCO. The output signal of the VCO is divided by the frequency divider and fed back to the phase detector. When the PLL frequency multiplier circuit is locked, the frequency of the VCO's output signal is equal to the product of the shaped pulse signal frequency and the frequency divider's division ratio. The frequency divider's division ratio is set to 4. The reason for setting the frequency divider's division ratio to 4 is that the pulse signal output by the crankshaft angle sensor generates four equally spaced pulses within each original pulse cycle after frequency multiplication. This increases the angular resolution of the angle synchronization signal to one-quarter of the original pulse angle resolution of the crankshaft angle sensor, providing more refined crankshaft angle position information for determining the equal angle increment sequence. The pulse signal output by the frequency multiplier circuit serves as the angle synchronization signal.

[0051] A detection system for cylinder deactivation camshaft displacement includes a memory, a processor, and a computer program stored in the memory and running on the processor. The memory is a random access memory or flash memory, storing a sequence of computer program instructions. The processor is a digital signal processor or a microcontroller. When the processor executes the computer program, it performs the following steps: acquiring the raw displacement signal output by the cylinder deactivation camshaft displacement sensor and simultaneously acquiring the angle synchronization signal output by the crankshaft angle sensor; performing empirical mode decomposition on the raw displacement signal to obtain an intrinsic mode function set and then resampling the intrinsic mode function set in the angular domain using the angle synchronization signal to obtain an angular domain mode function set; extracting displacement abrupt change features from the angular domain mode function set and converting these features into a displacement state symbol sequence; constructing a Markov transfer field matrix based on the displacement state symbol sequence and calculating the spectral radius of the Markov transfer field matrix; and comparing the spectral radius with a preset spectral radius threshold range and outputting the detection result of the camshaft displacement. The processor receives the conditioned digital signals from the voltage signal output by the cylinder deactivation camshaft displacement sensor and the pulse signal output by the crankshaft angle sensor through an input / output interface. The camshaft displacement detection results output by the processor are sent to the engine control unit via the communication interface.

[0052] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for detecting the displacement of a cylinder deactivation camshaft, characterized in that, include: Acquire the raw displacement signal output by the cylinder deactivation camshaft displacement sensor, and simultaneously acquire the rotation angle synchronization signal output by the crankshaft rotation angle sensor; The original displacement signal is subjected to empirical mode decomposition to obtain an intrinsic mode function set, and the intrinsic mode function set is resampled in the angular domain using the rotation synchronization signal to obtain an angular domain mode function set. Displacement abrupt change features are extracted from the angular domain mode function set, and the displacement abrupt change features are converted into a displacement state symbol sequence; A Markov transfer field matrix is ​​constructed based on the displacement state symbol sequence, and the spectral radius of the Markov transfer field matrix is ​​calculated. The spectral radius is compared with a preset spectral radius threshold range to output the detection result of the camshaft displacement.

2. The method for detecting the displacement of a cylinder-stopping camshaft according to claim 1, characterized in that, The process of performing empirical mode decomposition on the original displacement signal to obtain a set of intrinsic mode functions (EMFs), and then using the rotation synchronization signal to perform angular domain resampling on the EMF set to obtain an angular domain EMF set, specifically includes: The original displacement signal is subjected to extreme point detection, the upper and lower envelope mean values ​​are obtained by cubic spline interpolation, and the intrinsic mode functions of each order are iteratively screened to form the intrinsic mode function set. The angular synchronization signal is converted into an equal-angle increment sequence, and each intrinsic mode function in the intrinsic mode function set is interpolated and resampled according to the equal-angle increment sequence to generate the angular domain mode function set.

3. The method for detecting the displacement of a cylinder-stopping camshaft according to claim 1, characterized in that, The step of extracting displacement abrupt change features from the angular domain mode function set and converting the displacement abrupt change features into a displacement state symbol sequence specifically includes: Calculate the local energy density of each mode function in the angular domain mode function set, locate the angular domain interval where the local energy density exceeds the energy threshold, and use the amplitude change rate within the angular domain interval as the displacement abrupt change feature; The displacement abrupt change features are encoded according to the sign and numerical range of the amplitude change rate to generate the displacement state symbol sequence.

4. The method for detecting the displacement of a cylinder-stopping camshaft according to claim 1, characterized in that, The step of constructing the Markov transfer field matrix based on the displacement state symbol sequence and calculating the spectral radius of the Markov transfer field matrix specifically includes: The displacement state symbol sequence is divided into multiple time windows, and the transition probability between different symbols in each time window is calculated to form the Markov transition field matrix. The Markov transition field matrix is ​​decomposed into eigenvalues ​​to extract all eigenvalues, and the maximum value of the eigenvalue modulus is taken as the spectral radius.

5. The method for detecting the displacement of a cylinder-stopping camshaft according to claim 1, characterized in that, The step of comparing the spectral radius with a preset spectral radius threshold range and outputting the detection result of the camshaft displacement specifically includes: If the spectral radius falls within the preset spectral radius threshold range, a normal displacement signal is generated; If the spectral radius exceeds the preset spectral radius threshold range, a displacement anomaly signal is generated; Based on the normal displacement signal or the abnormal displacement signal, and combined with the maximum continuous symbol length in the displacement state symbol sequence, the detection result of the camshaft displacement is output.

6. The method for detecting the displacement of a cylinder-stopping camshaft according to claim 1, characterized in that, The process of acquiring the raw displacement signal output by the cylinder deactivation camshaft displacement sensor and simultaneously acquiring the angular synchronization signal output by the crankshaft angle sensor specifically includes: The voltage signal output by the cylinder deactivation camshaft displacement sensor and the pulse signal output by the crankshaft angle sensor are synchronously acquired at a preset sampling frequency. The voltage signal is amplified and filtered to obtain the original displacement signal; The pulse signal is shaped and frequency multiplied to obtain the angle synchronization signal.

7. The method for detecting the displacement of a cylinder-stopping camshaft according to claim 2, characterized in that, The step of interpolating and resampling each intrinsic mode function in the intrinsic mode function set based on the equal-angle increment sequence to generate the angular domain mode function set specifically includes: Align the starting point of the equal-angle increment sequence with the starting point of the intrinsic mode function set to determine the angular domain sampling time corresponding to each equal-angle increment; At the sampling time in the angular domain, the amplitude of each intrinsic mode function is interpolated using a piecewise cubic conformal interpolation algorithm to obtain the set of angular domain mode functions.

8. The method for detecting the displacement of a cylinder-stopping camshaft according to claim 3, characterized in that, The calculation of the local energy density of each mode function in the angular domain mode function set, and the location of angular domain intervals where the local energy density exceeds the energy threshold, with the amplitude change rate within the angular domain interval as the displacement abrupt change feature, specifically includes: The angular domain mode function set is divided into angular domain windows. The sum of squares of the magnitudes of all mode functions within each angular domain window is calculated, and the sum of squares is divided by the length of the angular domain window to obtain the local energy density. The local energy density is compared with the energy threshold, and a continuous angular domain window exceeding the energy threshold is retained as the angular domain interval; The ratio of the magnitude difference at the two endpoints of the angular domain interval to the angular domain span is calculated as the displacement abrupt change feature.

9. A method for detecting the displacement of a cylinder-stopping camshaft according to claim 4, characterized in that, The process of calculating the transition probabilities between different symbols within each time window to form the Markov transition field matrix specifically includes: Map each symbol in the displacement state symbol sequence to an index of the Markov transfer field matrix; For each time window, the frequency of adjacent symbol pairs is counted, and the frequency is divided by the total frequency of all adjacent symbol pairs within that time window to obtain the transition probability; Fill the corresponding positions of the Markov transition field matrix according to the order of the indexes to complete the construction of the Markov transition field matrix.

10. A detection system for cylinder deactivation camshaft displacement, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method for detecting the displacement of a cylinder-stopping camshaft as described in any one of claims 1 to 9.

Images