JP2010127264A - Control device of internal combustion engine - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a control device of an internal combustion engine, highly accurately determining a combustion state, in the control device of the internal combustion engine. <P>SOLUTION: A data clustering part 12 classifies the past data on the air-fuel ratio, the EGR ratio and a crank angle speed into two clusters, based on the specified two probability distributions. A separate hyperplane estimating part 14 specifies a boundary condition between a stable state and an unstable state by the past data. A system parameter estimating part 16 specifies the relationship valid between the past data in the predetermined time and the crank angle speed before by one step more than the time. A future combustion state is determined based on two estimating results being a result of respectively applying present data to the specified boundary condition and the relationship. Validity of a determining result is also evaluated by applying data including the estimated crank angle speed in the two probability distributions used when classifying the past data. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a control device for an internal combustion engine, and more particularly, to a control device for an internal combustion engine that determines a future combustion state of the internal combustion engine using rules or the like corresponding to the combustion state of the internal combustion engine.

  Conventionally, for example, in Patent Document 1, it is estimated when the average deviation between the current value and the previous value of the angular velocity detected within a predetermined crank angle range of the same cylinder and the ignition timing are advanced or retarded. A control device for an internal combustion engine that determines whether or not the combustion state of the internal combustion engine is in a stable state based on the difference between the average deviation between the current value and the previous value of the angular velocity is disclosed. In this control device, the combustion state is determined based on whether or not the difference between the average deviation of the actually measured values and the average deviation of the estimated values is within an allowable range.

JP 2003-286890 A Japanese translation of PCT publication No. 2002-527657

  By the way, in order to determine the combustion state of the internal combustion engine with high accuracy, it is desirable that an actual measurement value and an estimated value to be compared can be detected or estimated accurately. Patent Document 1 discloses a method for obtaining a next estimated value from the latest measured values using a least square method or a successive approximation method. However, the actual measurement value originally varies depending on the combustion state. For this reason, when the above estimated value is obtained from the actually measured value detected in the unstable combustion state, the accurate estimated value cannot always be estimated, and as a result, the combustion state of the internal combustion engine is highly accurate. In some cases, it could not be judged.

  In order to determine the combustion state with high accuracy, it is desirable that there are a plurality of determination indexes to be compared. In Patent Document 1, the combustion state is estimated from one determination index called an average deviation between the actual measurement value and the estimated value. For this reason, for example, if this one determination index is wrong, the determination result may be the opposite. Even in the case of Patent Document 1 in which the angular velocity may not be accurately estimated even if the determination index is not erroneous, the accuracy of combustion determination may be reduced if only one such determination index is used.

  An object of the present invention is to provide a control device for an internal combustion engine that can determine the combustion state with high accuracy.

In order to achieve the above object, a first invention is a control device for an internal combustion engine,
The unknown parameters of the first probability distribution corresponding to the combustion stable state of the internal combustion engine according to the past data of the exhaust gas recirculation rate, the air-fuel ratio and the crank angular velocity and the second probability distribution corresponding to the combustion unstable state of the internal combustion engine An unknown parameter is estimated by a maximum likelihood estimation method based on these data, and the probability distribution specifying means for specifying the first probability distribution and the second probability distribution by the estimated unknown parameter;
Based on the first probability distribution and the second probability distribution after specification, it is determined whether each of the past data is data of the combustion stable state or data of the combustion unstable state. Data classification means for classifying each into a combustion stable cluster and a combustion unstable cluster,
Boundary condition specifying means for estimating a boundary condition established at a boundary between the stable combustion state and the unstable combustion state based on the past data, and specifying the boundary condition based on the estimated unknown parameter;
An unknown parameter having a relationship established between exhaust gas recirculation rate, air-fuel ratio and crank angular velocity data at a predetermined time and crank angular velocity data at a time one step ahead of the predetermined time First rule specifying means for estimating based on stable cluster data, and specifying the relationship as a first rule corresponding to a stable combustion state by the estimated unknown parameter;
An unknown parameter having a relationship established between exhaust gas recirculation rate, air-fuel ratio and crank angular velocity data at a predetermined time and crank angular velocity data at a time one step ahead of the predetermined time A second rule specifying means for estimating based on stable cluster data and specifying the relationship as a second rule corresponding to the unstable combustion state by the estimated unknown parameter;
Data including the current values of the exhaust gas recirculation rate, air-fuel ratio, and crank angular velocity data are applied to the identified boundary conditions so that the data including the current values is classified into the stable state or the A category estimation means for estimating whether the category is classified into a stable state;
Crank angular velocity estimating means for estimating a future crank angular velocity by applying data including the current value to the first rule or the second rule;
Combustion state determination means for determining whether the future combustion state of the internal combustion engine is a stable state or an unstable state based on the estimation result by the section estimation unit and the future crank angular velocity;
Based on the first probability distribution and the second probability distribution after specification, it is determined whether the data including the current value is data of the combustion stable state or data of the combustion unstable state. And evaluating means for evaluating the validity of the determination result by the combustion state determining means.

The second invention is the first invention, wherein
Estimated fluctuation amount calculating means for calculating fluctuation amounts per predetermined time of a plurality of values of the future crank angular velocity as estimated fluctuation amounts;
A deviation calculating means for calculating a deviation between the estimated fluctuation amount of the actual crank angular speed of the internal combustion engine and the estimated fluctuation amount;
The combustion state determination means determines whether the future combustion state of the internal combustion engine is a stable state or an unstable state depending on whether the deviation is less than a predetermined first determination value. Features.

  According to the first invention, the future crank angular velocity can be estimated from the two rules, the first rule corresponding to the stable combustion state and the second rule corresponding to the unstable combustion state. These two rules are based on two probability distributions followed by past data composed of EGR rate, air-fuel ratio and crank angular velocity, ie, a first probability distribution corresponding to the combustion stable state and a second probability distribution corresponding to the combustion unstable state. These past data are classified into a probability distribution and specified based on a combustion stable cluster and a combustion unstable cluster which are data groups after the classification. Therefore, according to the first invention, the future crank angular speed can be estimated with the two rules specified as described above, and therefore, the future crank angular speed can be more accurately compared with the case of estimating using a single model. Can be estimated.

  Furthermore, according to the first aspect of the invention, the data including the current values of the exhaust gas recirculation rate, the air-fuel ratio, and the crank angular velocity is stabilized by the boundary condition established at the boundary between the stable combustion state and the unstable combustion state. It can be estimated whether it is classified into a state or an unstable state. Then, by combining the classification based on the boundary condition and the two rules described above, it is possible to switch the rule for estimating the future crank angular speed depending on which state the data including the current value is in. Become. Therefore, according to the first aspect, it is possible to estimate in which state the data including the current value is, and to estimate the future crank angular velocity at that time with high accuracy.

  Further, according to the first invention, the future combustion state of the internal combustion engine can be determined based on the estimation result by the segment estimation means and the future crank angular speed by the crank angular speed estimation means. Therefore, it is possible to determine the future combustion state of the internal combustion engine using these two estimation results as the determination index for combustion determination.

  In addition, according to the first invention, based on the first probability distribution and the second probability distribution after specification, the data including the current value is data of the combustion stable state, or It is possible to determine whether the data is an unstable combustion state data. Then, it is possible to evaluate the credibility of the current value data used when determining the combustion state. Therefore, the validity of the determination result of the future combustion state can be evaluated.

  According to the second aspect of the present invention, the fluctuation amount calculated based on the estimated future crank angular velocity is compared with the fluctuation amount based on the actual crank angular velocity, and the future combustion state is determined using these deviations as a determination index for combustion determination. be able to.

Embodiment 1. FIG.
The control device of the first embodiment is configured inside an ECU (Electronic Control Unit) mounted on a vehicle or the like. For this reason, first, the control apparatus which concerns on this embodiment is demonstrated using FIG. FIG. 1 is a functional block diagram when the ECU functions as a control device that determines the combustion state of the internal combustion engine.

  As shown in FIG. 1, the ECU includes a system identification unit 10, a crank angular speed estimation unit 18, a crank angular speed fluctuation amount calculation unit 20, an attribution probability calculation unit 22, and an abnormality determination unit 24. The system identification unit 10 includes a data clustering unit 12, a separated hyperplane estimation unit 14, and a system parameter estimation unit 16.

(System identification unit 10)
The system identification unit 10 uses a piecewise affine self in which an identification expression describing engine characteristics from an air-fuel ratio, an EGR rate to a crank angular velocity (rotational behavior) is switched between a combustion stable state and a combustion unstable state of the internal combustion engine. Identification is performed as a regression (Piece-Wise affine Auto-Regressive eXogeneous; PWARX) model. The system identification unit 10 determines parameters θ ₁ , θ ₂ , a, and b of equations (3) and (4) described later based on past time series data, and identifies the PWARX model. To do.

  The identification is performed by using the past actual air-fuel ratio, EGR rate, and crank angular velocity as past time-series data. Let the air-fuel ratio at time k be λ (k), the EGR rate be e (k), and the crank angular speed be ω (k). If time k is the current time, the past time series data is λ (k-1), λ (k-2), ..., e (k-1), e (k-2), ... , Ω (k-1), ω (k-2),. These past time-series data are assumed to be buffered in the ECU.

上式（１）が燃焼安定状態に対応し、上式（２）が燃焼不安定状態に対応するものとする。上式（１）及び（２）において、ε(k)は式誤差である。上式（１）、（２）は、それぞれ次式（３）、（４）のように書き表すことができる。

上式（３）、（４）において、右肩の添え字「T」は転置行列を意味する。また、x(k)は回帰ベクトルであり、過去の空燃比、ＥＧＲ率及びクランク角速度を用いて、
x(k)=[ω(k-1) ・・・ ω(k-n) λ(k-1) ・・・ λ(k-m) e(k-1) ・・・ e(k-l)）]^T ・・・（５）
により与えられる。上式（５）において、kは、k=1,2,・・・,Nの離散時刻の値であり、Nはデータ数である。また、自然数n,m,lは、それぞれモデルを記述するために必要な過去の時系列データの個数を表す。
また、上式（３）、（４）において、θ₁及びθ₂は、未知のパラメータであり、次式（６）、（７）により与えられる。

Assuming that engine characteristics based on past time-series data follow the PWARX model, the relationship among the air-fuel ratio, EGR rate, and crank angular velocity is expressed by the following equations (1) and (2).

The above equation (1) corresponds to the stable combustion state, and the above equation (2) corresponds to the unstable combustion state. In the above equations (1) and (2), ε (k) is an equation error. The above formulas (1) and (2) can be written as the following formulas (3) and (4), respectively.

In the above formulas (3) and (4), the superscript “T” on the right shoulder means a transposed matrix. Also, x (k) is a regression vector, using the past air-fuel ratio, EGR rate and crank angular velocity,
x (k) = [ω (k-1) ... ω (kn) λ (k-1) ... λ (km) e (k-1) ... e (kl))] ^T・ ・・ (5)
Given by. In the above equation (5), k is a discrete time value of k = 1, 2,..., N, and N is the number of data. The natural numbers n, m, and l represent the number of past time series data necessary for describing the model.
In the above equations (3) and (4), θ ₁ and θ ₂ are unknown parameters and are given by the following equations (6) and (7).

Moreover, as shown in the above formulas (3) and (4), the combustion stable state and the combustion unstable state are assumed to be divided by a separation hyperplane on the regression vector space. The separation hyperplane is a boundary between these two states, and is expressed by the following equation (8).
a ^T x (k) + b = 0 (8)
In the above equation (8), a and b are unknown parameters that define the separation hyperplane, and are given by the following equation (9).

(Data clustering unit 12)
A data clustering unit 12 in the system identification unit 10 classifies past time series data into a combustion stable state cluster C ₁ corresponding to a combustion stable state of the internal combustion engine or a combustion unstable cluster C ₂ corresponding to a combustion unstable state. Is configured to do.

In the classification, first, an observation data vector z _k = (x ^T (k) ω ^T (k)) ^T is defined. Then, it is assumed that N pieces of observation data z ₁ ,..., Z _N of the observation data vector z _k follow a mixed normal distribution composed of normal distributions corresponding to the combustion stable state and the combustion unstable state.

上式（１０）において、α₁、α₂は、混合正規分布の混合割合であり、α₁，α₂≧0，α₁+α₂=1の関係を満たす。また上式（１０）〜（１２）において、μ₁，μ₂は、燃焼安定状態、燃焼不安定状態のそれぞれに対応する正規分布の平均ベクトルを示す。同様に、Σ₁、Σ₂は、それぞれに対応する正規分布の共分散行列を示す。添え字「n_z」は、観測データベクトルz_kの次元数を示す。さらにdetΣは、Σの行列式を示し、行列の右肩の添え字「-1」は逆行列を意味する。 The mixed normal distribution is expressed by the following equation (10) using the parameter Φ = (α ₁ , α ₂ , μ ₁ , μ ₂ , Σ ₁ , Σ ₂ ).

In the above equation (10), α ₁ and α ₂ are the mixing ratio of the mixed normal distribution, and satisfy the relationship of α ₁ , α ₂ ≧ 0, α ₁ + α ₂ = 1. In the above formulas (10) to (12), μ ₁ and μ ₂ represent average vectors of normal distributions corresponding to the combustion stable state and the combustion unstable state, respectively. Similarly, Σ ₁ and Σ ₂ indicate normal distribution covariance matrices corresponding to each. The subscript “n _z ” indicates the number of dimensions of the observation data vector z _k . Furthermore, detΣ indicates a determinant of Σ, and the subscript “−1” on the right shoulder of the matrix means an inverse matrix.

を最大化するパラメータΦは、ＥＭアルゴリズムを用いることにより推定することができる。ＥＭアルゴリズムでは、パラメータΦを反復的に計算することができる。 Subsequently, the parameter Φ of the mixed normal distribution is obtained by the maximum likelihood estimation method. Likelihood function

The parameter Φ that maximizes can be estimated by using the EM algorithm. In the EM algorithm, the parameter Φ can be calculated iteratively.

Subsequently, the observed data is classified using the obtained parameter Φ. Here, the assigned probability that the observation data belongs to the combustion stable state cluster C ₁ is α ₁ p ₁ (z _k ; Φ), and the belonging probability that belongs to the combustion unstable state cluster C ₂ is α ₂ p ₂ (z _k ; Φ). The classification is performed into clusters corresponding to partial distributions having higher attribution probabilities. That is, data of α ₁ p ₁ (z _k ; Φ) ≧ α ₂ p ₂ (z _k ; Φ) is stored in the combustion stable state cluster C ₁ and α ₁ p ₁ (z _k ; Φ) <α ₂ p ₂ ( The data of z _k ; Φ) is classified into the combustion unstable cluster C ₂ . By doing so, the times k = 1, 2,..., N corresponding to the observation data can be classified into the respective clusters.

(Separated hyperplane estimation unit 14)
The separation hyperplane estimation unit 14 in the system identification unit 10 is configured to estimate the parameters of the separation hyperplane on the regression space that divides the combustion stable state and the combustion unstable state. The parameters a and b of the separation hyperplane given by the above equation (8) can be estimated by using a soft margin support vector machine.

  A support vector machine is an algorithm that can determine which of two classes a given data belongs to, and that is true when there is a correct class for all sample data (linear separation is possible). This is a method of adjusting the parameters a and b of the plane. However, there is no single hyperplane that realizes such linear separation. Therefore, the distance from the hyperplane to each class is defined as a margin (1 / || a ||), and the hyperplane parameters a and b are determined so as to maximize this margin.

このように推定されたパラメータa,bは、ＥＣＵに記憶される。 Here, a correct class does not always exist for all sample data (not linearly separable). In such a case, a soft margin support vector machine in which a classification error term is introduced into the support vector machine is used. That is, when the data enters the opposite class beyond the hyperplane, the hyperplane parameters a and b are adjusted so as to minimize the sum of the distances v exceeding the hyperplane. In summary, if a soft margin support vector machine is used, the parameters a and b can be estimated by solving the quadratic optimization problem shown in the following equation (13).

The parameters a and b estimated in this way are stored in the ECU.

上式（１４）及び（１５）において、N₁,N₂は、それぞれのクラスタC₁,C₂に含まれるデータ数を示す。このように推定されたパラメータθ₁^,θ₂^は、ＥＣＵに記憶される。 (System parameter estimation unit 16)
The system parameter estimation unit 16 in the system identification unit 10 is configured to estimate the parameters θ ₁ and θ ₂ using data corresponding to discrete times classified by the data clustering unit 12. The parameters θ ₁ and θ ₂ can be estimated by the following equations (14) and (15).

In the above equations (14) and (15), N ₁ and N ₂ indicate the number of data included in the respective clusters C ₁ and C ₂ . The parameters θ ₁ ^ and θ ₂ ^ estimated in this way are stored in the ECU.

尚、上式（１６）が本発明の第１規則に、上式（１７）が本発明の第２規則にそれぞれ該当する。また、上式（１６）及び（１７）において、x(k+1)は回帰ベクトルであり、時刻kまでの空燃比、ＥＧＲ率及びクランク角速度を用いて、
x(k+1)=[ω(k) ・・・ ω(k-n+1) λ(k) ・・・ λ(k-m+1) e(k) ・・・ e(k-l+1)]^T ・・・（１８）
により与えられることになる。 As described above, the system identification unit 10 identifies the PWARX model in which the identification formula is switched between the combustion stable state and the combustion unstable state of the internal combustion engine. In summary, assuming that the current time is time k, the estimated value ω ^ (k + 1) of the crank angular speed at time k + 1 is based on the data history of the air-fuel ratio, EGR rate, and crank angular speed up to time k−1. Using the parameters θ ₁ ^ and θ ₂ ^ estimated in this way, the following equations (16) and (17) are used. The subscript “^” used here represents an estimated value.

The above equation (16) corresponds to the first rule of the present invention, and the above equation (17) corresponds to the second rule of the present invention. In the above equations (16) and (17), x (k + 1) is a regression vector, and uses the air-fuel ratio, EGR rate and crank angular speed up to time k,
x (k + 1) = [ω (k) ... ω (k-n + 1) λ (k) ... λ (k-m + 1) e (k) ... e (k-l +1)] ^T ... (18)
Will be given by.

  As described above, the estimated value ω ^ (k + 1) of the crank angular speed at time k + 1 is assigned to the air / fuel ratio, EGR rate and crank angular speed data history up to time k-1, and the current crank angular speed data. Can be done each time. Therefore, there is an effect that the degree of freedom of model identification is high.

(Crank angular velocity estimation unit 18)
The crank angular velocity estimation unit 18 uses the parameters a and b and the above equation (18) to estimate whether the combustion state of the internal combustion engine is stable or unstable, and the estimation result and the above equation (16 ) Or (17) is used to estimate ω ^ (k + 1).

Specifically, first, the parameters a and b estimated from the data history up to time k-1, and the regression vector x () generated by adding ω (k) which is crank angular velocity data at time k to this data history. a ^T x (k + 1) + b is calculated from (k + 1). If a ^T x (k + 1) + b ≧ 0, the crank angular velocity estimation unit 18 estimates that the internal combustion engine is in a stable combustion state, and the combustion state prediction flag i ^ (k + 1) = 1 is set. Buffered in the ECU. On the other hand, if a ^T x (k + 1) + b <0, the crank angular velocity estimation unit 18 estimates that the internal combustion engine is in an unstable combustion state, and the combustion state prediction flag i ^ (k + 1) = 0 is set to ECU. Buffered. Subsequently, the estimated value ω ^ (k + 1) of the crank angular velocity at time k + 1 is estimated using the combustion state prediction flag value and the above equation (16) or (17). The estimated ω ^ (k + 1) is buffered in the ECU.

(Crank angular velocity fluctuation calculation unit 20)
The crank angular speed fluctuation calculating unit 20 is configured to calculate a deviation between the estimated crank angular speed fluctuation and the measured crank angular speed fluctuation. Examples of the deviation of the variation amount of the crank angular speed include a time difference of the crank angular speed. Further, the time difference of the crank angular speed is not limited to the difference between the previous value and the current value of the crank angular speed, and an average deviation of a plurality of crank angular speeds can be applied. As the difference between the previous value and the current value of the crank angular speed, for example, the estimated value of the crank angular speed ω ^ (k + 1) estimated at time k + 1 and the estimated value of the crank angular speed ω ^ ( k) Δω ^ = ω ^ (k + 1) -ω ^ (k + 2), crank angular velocity ω (k + 1) at time k + 1 detected by the crank position sensor, and time k + 2 The difference Δω = ω (k + 1) −ω (k + 2) with respect to the crank angular speed ω (k + 2) in FIG. Δω ^ −Δω is calculated from Δω ^ and Δω thus obtained. The calculated Δω ^ −Δω is buffered in the ECU.

(Attribution probability calculation unit 22)
The attribution probability calculating unit 22 uses the parameter Φ and the above equation (11) or (12), and the attribution probability α ₁ p ₁ (z where the observation data vector z _{k at the} time k belongs to the combustion stable state cluster C _{1. k} ; Φ) and the belonging probability α ₂ p ₂ (z _k ; Φ) attributed to the combustion unstable cluster C ₂ . As described above, the data clustering unit 12 estimates the parameter Φ based on the time series data up to the time k−1.

Membership probability, first, similar observation data vector z _k defined cluster analysis unit 12, the observation data vector ^{_{z k = (x T (k}} ) ω T (k)) that defines the ^T. Then, α ₁ p ₁ (z _k ; Φ) and α ₂ p ₂ (z _k ; Φ) are calculated using the parameter Φ estimated by the data clustering unit 12. Then, α ₁ p ₁ (z _k ; Φ) and α ₂ p ₂ (z _k ; Φ) after the calculation are buffered in the ECU.

In the present embodiment, the crank angular velocity is estimated a plurality of times when calculating Δω. Therefore, for example, when estimated crank angular velocity values ω ^ (k + 1) and ω ^ (k + 2) are estimated at time k + 1 and time k + 2, the associated probability α ₁ p ₁ (z _k ; Φ), α ₂ p ₂ (z _k ; Φ) and α ₁ p ₁ (z _{k + 1} ; Φ), α ₂ p ₂ (z _{k + 1} ; Φ) 1 and time k + 2 are respectively calculated and buffered in the ECU.

(Abnormality determination unit 24)
The abnormality determination unit 24 includes a combustion state prediction flag i ^ (k + 1), a deviation Δω ^ -Δω from the fluctuation amount of the crank angular velocity, an attribution probability α ₁ p ₁ (z _k ; Φ) and α ₂ p ₂ (z _k ; Φ) is used to determine whether or not the combustion state of the internal combustion engine is abnormal. Abnormality determination includes (i) a result of the combustion state prediction flag i ^ (k + 1) value, (ii) a comparison result between a deviation Δω ^ -Δω and a predetermined deviation from the fluctuation amount of the crank angular speed, and (iii) α _This is done by using the comparison result between ₁ p ₁ (z _k ; Φ) and α ₂ p ₂ (z _k ; Φ).

  In (i) above, it is determined whether the combustion state prediction flag i ^ (k + 1) estimated by the crank angular velocity estimation unit 18 is in a stable combustion state. That is, it is determined whether i ^ (k + 1) = 1. When i ^ (k + 1) = 0, there is a possibility of abnormal combustion. The abnormality determination unit 24 is configured to generate an abnormality signal when i ^ (k + 1) = 0.

  In (ii) above, the deviation Δω ^ −Δω from the crank angular speed variation calculated by the crank angular speed variation calculator 20 is compared with a predetermined set value (threshold). This threshold value can be freely set according to the required degree of determination accuracy of the combustion state. If Δω ^ −Δω is equal to or greater than the threshold value, there is a discrepancy between the estimated variation amount of the crank angular velocity and the actually measured variation amount of the crank angular velocity. For this reason, if the deviation between ω ^ -Δω and the threshold value is greater than or equal to a certain value, the possibility of an abnormal combustion state is recognized. The abnormality determination unit 24 is configured to generate an abnormality signal when the deviation between ω ^ −Δω and the threshold value is equal to or greater than a certain value.

In the above (iii), the attribution probability α ₁ p ₁ (z _k ; Φ) is compared with α ₂ p ₂ (z _k ; Φ). And, when the belonging probability α ₁ p ₁ (z _k ; Φ) ≧ α ₂ p ₂ (z _k ; Φ), the combustion stability probability is high, but α ₁ p ₁ (z _k ; Φ) <α ₂ In the case of p ₂ (z _k ; Φ), the probability of combustion instability is high. In other words, the observation data used when the combustion instability probability is high in (iii) even though the combustion stability in (i) above is determined, i.e., i ^ (k + 1) = 1. It is recognized that the vector z _k may have low credibility. The abnormality determination unit 24 is configured to generate an abnormality signal when α ₁ p ₁ (z _k ; Φ) <α ₂ p ₂ (z _k ; Φ).

  The abnormality determination is made based on whether or not the abnormality signals (i) to (iii) are detected a set number of times within a predetermined determination time. Specifically, when the abnormality signals (i) to (iii) are detected, the counter value of the abnormality counter built in the abnormality determination unit 24 is increased. Then, when the counter value reaches a set value or more within a predetermined determination time, it is determined that the combustion state of the internal combustion engine is abnormal.

[Specific Processing in Embodiment 1]
Next, a combustion state determination routine executed by the ECU in the present embodiment will be described with reference to FIG. The routine in FIG. 2 is repeated for a determination time period for performing the combustion state determination, for example, every crank angular velocity acquisition timing from the crank position sensor.

First, in step 100, model parameters used in the present embodiment are read. As described above, the system identification unit 10 estimates the parameters θ ₁ ^ and θ ₂ ^ of the PWARX model based on the time series data up to time k−1, and these are stored in the ECU. Similarly, the separation hyperplane estimation unit 14 estimates the parameters a and b of the separation hyperplane, and these are stored in the ECU. For this reason, these parameters are read into the ECU. Subsequent to step 100, in step 120, time-series data of the air-fuel ratio λ, the EGR rate, and the crank angular speed up to time k is read.

Subsequently, in step 140, a ^T x (k + 1) + b is calculated from the parameters a and b read in step 100 and the regression vector x (k + 1). In this step, a regression vector x (k + 1) is generated using the parameters a and b read in step 100 and the time series data up to time k read in step 120, and then the above a ^T x ( k + 1) + b is calculated. If a ^T x (k + 1) + b ≧ 0, i ^ (k + 1) = 1, and if i ^ (k + 1) <0, i ^ (k + 1) = 0 (Step 160).

  Subsequently, in step 180, it is determined whether or not the combustion is stable. If i ^ (k + 1) = 1 buffered at step 160, the routine proceeds to step 200, where an estimated value ω ^ (k + 1) of the crank angular velocity is calculated. This ω ^ (k + 1) is buffered in the ECU at step 220. On the other hand, if i ^ (k + 1) = 1 is not satisfied at step 180, the routine proceeds to step 240, where the estimated crank angular velocity value ω ^ (k + 1) is calculated. Buffered.

  Subsequent to step 220 or 260, in step 280, the crank angular speed ω (k) at time k, the estimated crank angular speed ω ^ (k), and the crank angular speed ω (k + 1) at time k + 1 are read. Subsequently, in step 300, Δω ^ −Δω is calculated based on these values. As described above, Δω ^ −Δω is calculated by the crank angular velocity estimation unit 18. The calculated Δω ^ −Δω is buffered in the ECU at step 320.

Subsequently, in step 340, the attribution probabilities α ₁ p ₁ (z _k ; Φ) and α ₂ p ₂ (z _k ; Φ) are calculated. As described above, in the membership probability calculation unit 22, the observation data vector z _k at time k is the combustion attributed to the stable state cluster C ₁ membership probability _{α 1 p 1 (z k;} Φ) and combustion instability cluster C ₂ The membership probability α ₂ p ₂ (z _k ; Φ) belonging to is calculated. The attribution probabilities α ₁ p ₁ (z _k ; Φ) and α ₂ p ₂ (z _k ; Φ) are buffered in the ECU in step 360. Subsequent to step 360, in step 380, the combustion state of the internal combustion engine is determined.

  Next, the combustion state determination subroutine executed in step 380 will be described with reference to the flowchart of FIG. First, in step 400, it is determined whether or not an abnormality counter built in the abnormality determination unit 24 is ON. If it is determined that it is not ON, the abnormality counter is turned ON in step 420. On the other hand, if it is determined that the abnormality counter is ON, the process proceeds to step 440.

  In step 440, it is determined whether or not the combustion state prediction flag i ^ (k + 1) = 1. As described above, the combustion state prediction flag i ^ (k + 1) is buffered in the ECU through the process of step 160. Therefore, it is determined whether or not the data at the time k + 1 is in the combustion stable state by determining whether or not the combustion state prediction flag i ^ (k + 1) = 1. If i ^ (k + 1) = 1, go to Step 480. If not, the process proceeds to step 460, and the counter value of the abnormality counter is incremented by one.

  In step 480, the deviation Δω ^ −Δω from the fluctuation amount of the crank angular velocity is compared with a predetermined set value (threshold value). As described above, Δω ^ −Δω is buffered in the ECU through the process of step 300. The threshold value is stored separately in the ECU. Therefore, this Δω ^ −Δω is compared with a predetermined threshold value. If | Δω ^ −Δω | <threshold, the process proceeds to step 500. If not, the process proceeds to step 460, and the counter value of the abnormality counter is incremented by one.

In step 500, a comparison is made between attribution probabilities α ₁ p ₁ (z _k ; Φ) and α ₂ p ₂ (z _k ; Φ). As described above, the attribution probabilities α ₁ p ₁ (z _k ; Φ) and α ₂ p ₂ (z _k ; Φ) are buffered in the ECU through the process of step 360. For this reason, in step 500, a comparison is made between these attribution probabilities. If α ₁ p ₁ (z _k ; Φ) ≧ α ₂ p ₂ (z _k ; Φ), the routine proceeds to step 520, where the abnormality counter is reset. If not, the process proceeds to step 460, and the counter value of the abnormality counter is incremented by one.

  In step 540, it is determined whether or not the counter value of the abnormality counter is equal to or greater than a predetermined set value (threshold value). As described above, the abnormality determination unit 24 determines that the combustion state of the internal combustion engine is abnormal, that is, the combustion unstable state, when the abnormality counter reaches a set value or more within a predetermined determination time. Therefore, in step 540, the counter value is compared with the threshold value. If the counter value is equal to or greater than the threshold value, the process proceeds to step 580 and is determined to be abnormal. Otherwise, the process proceeds to step 560 and is determined to be normal.

  As described above, according to the routines shown in FIGS. 2 and 3, the future crank angular speed can be sequentially estimated using the PWARX model identified based on the past time series data. Further, the deviation Δω ^ −Δω of the variation amount of the crank angular speed can be sequentially estimated based on the estimated crank angular speed. Further, it is possible to predict the future combustion state using the parameters of the separated hyperplane estimated when the PWARX model is identified. It is also possible to calculate statistical information used in identifying the PWARX model. By using these three diagnostic elements, the combustion state of the internal combustion engine can be determined with high accuracy.

In the above-described embodiment, the data clustering unit 12 estimates the parameter Φ of the mixed normal distribution and specifies the above equation (10) so that the “probability distribution specifying means” in the first invention is the data By classifying the observation data in the clustering unit 12, the “data classification unit” in the first invention estimates the parameters a and b of the separation hyperplane in the separation hyperplane estimation unit 14, and the above equation (8 ) Is specified, the “boundary condition specifying means” in the first aspect of the invention estimates the parameters θ ₁ ^, θ ₂ ^ in the system parameter estimation unit 16 and specifies the above equations (16) and (17) Thus, the “first rule specifying means” and the “second rule specifying means” in the first invention are realized. Further, when the ECU executes the process of step 140, the “segment estimation means” in the first aspect of the invention is executed. When the ECU executes the process of step 200 or 240, the “crank angular velocity” of the first aspect of the invention is set. When the “estimating means” executes the processing of steps 440 to 480 and 520, the “combustion state determination means” in the first aspect of the invention executes the processing of steps 500, 460 and 520. Each of the “evaluation means” in the present invention is realized.

  Further, in the above-described embodiment, the crank angular speed fluctuation amount calculation unit 20 calculates the time difference of the estimated crank angular speed, whereby the “estimated fluctuation amount calculation means” in the second aspect of the invention is determined by the ECU in step 300. By executing the process, the “deviation calculating means” in the second aspect of the invention is realized.

It is a functional block diagram in case ECU functions as a control apparatus which performs the combustion state determination of an internal combustion engine. It is a flowchart which shows the combustion state determination routine which ECU performs. It is a flowchart which shows the combustion state determination subroutine which ECU performs.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 System identification part 12 Data clustering part 14 Separation hyperplane estimation part 16 System parameter estimation part 18 Crank angular speed estimation part 20 Crank angular speed fluctuation amount calculation part 22 Attribution probability calculation part 24 Abnormality determination part

Claims (2)

The unknown parameters of the first probability distribution corresponding to the combustion stable state of the internal combustion engine according to the past data of the exhaust gas recirculation rate, the air-fuel ratio and the crank angular velocity and the second probability distribution corresponding to the combustion unstable state of the internal combustion engine An unknown parameter is estimated by a maximum likelihood estimation method based on these data, and the probability distribution specifying means for specifying the first probability distribution and the second probability distribution by the estimated unknown parameter;
Based on the first probability distribution and the second probability distribution after specification, it is determined whether each of the past data is data of the combustion stable state or data of the combustion unstable state. Data classification means for classifying each into a combustion stable cluster and a combustion unstable cluster,
Boundary condition specifying means for estimating a boundary condition established at a boundary between the stable combustion state and the unstable combustion state based on the past data, and specifying the boundary condition based on the estimated unknown parameter;
An unknown parameter having a relationship established between exhaust gas recirculation rate, air-fuel ratio and crank angular velocity data at a predetermined time and crank angular velocity data at a time one step ahead of the predetermined time First rule specifying means for estimating based on stable cluster data, and specifying the relationship as a first rule corresponding to a stable combustion state by the estimated unknown parameter;
An unknown parameter having a relationship established between exhaust gas recirculation rate, air-fuel ratio and crank angular velocity data at a predetermined time and crank angular velocity data at a time one step ahead of the predetermined time A second rule specifying means for estimating based on stable cluster data and specifying the relationship as a second rule corresponding to the unstable combustion state by the estimated unknown parameter;
Data including the current values of the exhaust gas recirculation rate, air-fuel ratio, and crank angular velocity data are applied to the identified boundary conditions so that the data including the current values is classified into the stable state or the A category estimation means for estimating whether the category is classified into a stable state;
Crank angular velocity estimating means for estimating a future crank angular velocity by applying data including the current value to the first rule or the second rule;
Combustion state determination means for determining whether the future combustion state of the internal combustion engine is a stable state or an unstable state based on the estimation result by the section estimation unit and the future crank angular velocity;
Based on the first probability distribution and the second probability distribution after specification, it is determined whether the data including the current value is data of the combustion stable state or data of the combustion unstable state. Performing evaluation means for evaluating the validity of the determination result by the combustion state determination means;
A control device for an internal combustion engine, comprising: Estimated fluctuation amount calculating means for calculating fluctuation amounts per predetermined time of a plurality of values of the future crank angular velocity as estimated fluctuation amounts;
A deviation calculating means for calculating a deviation between the estimated fluctuation amount of the actual crank angular speed of the internal combustion engine and the estimated fluctuation amount;
The combustion state determination means determines whether the future combustion state of the internal combustion engine is a stable state or an unstable state depending on whether the deviation is less than a predetermined first determination value. The control apparatus for an internal combustion engine according to claim 1, wherein the control apparatus is an internal combustion engine.

Images