CN113915017A

CN113915017A - Engine characteristic estimation device and method, state estimation device, and storage medium

Info

Publication number: CN113915017A
Application number: CN202110777806.1A
Authority: CN
Inventors: 川谷圣; 藤原真; 笹岛己喜朗; 福岛圭一郎; 奥列克西·邦达连科; 北川泰士
Original assignee: National Research And Development Corp Marine Harbor Aviation Technology Research Institute; Nabtesco Corp
Current assignee: National Research And Development Corp Marine Harbor Aviation Technology Research Institute; Nabtesco Corp; National Institute of Maritime Port and Aviation Technology
Priority date: 2020-07-10
Filing date: 2021-07-09
Publication date: 2022-01-11
Also published as: JP2022015997A; KR20220007552A

Abstract

Provided are an engine characteristic estimation device and method, a state estimation device, and a storage medium, which can estimate the change of the engine characteristic with high precision. An engine characteristic estimation device (100) is provided with: a first calculation unit that calculates a first engine state parameter (X1) based on a first calculation model and a fuel supply amount to an engine (200); a second calculation unit that calculates a second engine state parameter (X2) on the basis of a second calculation model and the operation data of the engine (200); and an engine characteristic estimation section (130) that estimates a characteristic of the engine (200) based on the first engine state parameter (X1) and the second engine state parameter (X2).

Description

Engine characteristic estimation device and method, state estimation device, and storage medium

Technical Field

The present invention relates to a characteristic estimation technique and a state estimation technique of an engine.

Background

Engines are widely used in ships, automobiles, aircrafts, and the like, but due to increased awareness of environmental issues, further improvement in efficiency has been desired in recent years. In order to improve the efficiency of the engine, it is necessary to estimate the state of the engine with high accuracy and to optimally control the engine based on the estimation result.

Documents of the prior art

Patent document

Patent document 1: japanese Kokai publication No. 2009-510327

Patent document 2: japanese patent laid-open publication No. 2015-222074

Disclosure of Invention

Problems to be solved by the invention

Patent document 1 discloses a technique for estimating the state of an engine by a state observer. The state observer calculates a parameter indicating a state of the engine from various measurement data of the engine using a matrix indicating characteristics of the engine. In this calculation, when the matrix accurately represents the characteristics of the engine, the state parameters of the engine can be calculated with high accuracy. On the other hand, when the characteristics of the engine change due to aging degradation or a change in the external environment such as the intake air temperature, the matrix does not accurately reflect the characteristics of the engine, and therefore the estimation accuracy of the state of the engine decreases.

The present invention has been made in view of such circumstances, and an object thereof is to provide an engine characteristic estimation device capable of accurately estimating the characteristics of an engine even when the characteristics of the engine have changed. Another object of the present invention is to provide an engine state estimation device capable of estimating the state of an engine with high accuracy.

Means for solving the problems

In order to solve the above problem, an engine characteristic estimation device according to an aspect of the present invention includes: a first calculation unit that calculates a first engine state parameter that is a state variable of an engine, based on a first calculation model that represents a characteristic of the engine and a fuel supply amount that is supplied to the engine at an arbitrary time point; a second calculation unit that calculates a second engine state parameter that is a state variable of the engine, based on a second calculation model that is different from the first calculation model and that represents a characteristic of the engine, and on operation data relating to an operation of the engine that is driven at an arbitrary time point; and an engine characteristic estimation unit that estimates a characteristic of the engine at an arbitrary time point based on the first engine state parameter and the second engine state parameter.

In this aspect, the fuel supply amount used for the calculation by the first calculation unit is data that is not affected by changes in the characteristics of the engine due to changes in the external environment such as aging degradation and intake air temperature, and therefore the first engine state parameter that is the calculation result thereof is not affected by changes in the characteristics of the engine. In contrast, since the engine operation data used for the calculation by the second calculation unit is data affected by the change in the engine characteristic, the second engine state parameter that is the calculation result is affected by the change in the engine characteristic. In this way, by using two types of engine state parameters having different influences of the engine characteristic variation, the engine characteristic estimating section can estimate the engine characteristic variation with high accuracy. Further, even when the characteristics of the engine do not change, the state of the engine can be estimated with high accuracy by using two calculation units having different calculation models at the same time.

Another embodiment of the present invention is an engine characteristic estimation method. The method comprises the following steps: a first calculation step of calculating a first engine state parameter as a state variable of an engine based on a first calculation model representing a characteristic of the engine and a fuel supply amount supplied to the engine at an arbitrary time point; a second calculation step of calculating a second engine state parameter as a state variable of the engine based on a second calculation model different from the first calculation model representing a characteristic of the engine and action data on an action of the engine driven at an arbitrary time point; and an engine characteristic estimating step of estimating a characteristic of the engine at an arbitrary point in time based on the first engine state parameter and the second engine state parameter.

Another embodiment of the present invention is an engine state estimation device. The device is provided with: a first calculation unit that calculates a first engine state parameter that is a state variable of an engine, based on a first calculation model that represents a characteristic of the engine and a fuel supply amount that is supplied to the engine at an arbitrary time point; a second calculation unit that calculates a second engine state parameter that is a state variable of the engine, based on a second calculation model that is different from the first calculation model and that represents a characteristic of the engine, and on operation data relating to an operation of the engine that is driven at an arbitrary time point; and an engine state estimation unit that estimates the state of the engine at an arbitrary time point based on the first engine state parameter and the second engine state parameter.

According to this aspect, the state of the engine can be estimated with high accuracy by using two calculation units having different calculation models at the same time.

In addition, an arbitrary combination of the above-described constituent elements, and a mode in which the expression of the present invention is converted between a method, an apparatus, a system, a recording medium, a computer program, and the like are also effective as modes of the present invention.

ADVANTAGEOUS EFFECTS OF INVENTION

According to the present invention, even when the characteristics of the engine change, the characteristics of the engine can be estimated with high accuracy. In addition, the state of the engine can be estimated with high accuracy.

Drawings

Fig. 1 is a diagram showing the overall configuration of an engine characteristic estimation device according to an embodiment.

Fig. 2 is a diagram showing the structure of the engine.

Fig. 3 is a diagram showing the structure of another engine.

Fig. 4 is a diagram showing the structures of the simulator and the observer.

Fig. 5 is a diagram showing the structure of the engine characteristic estimating unit.

Fig. 6 is a flowchart showing a flow of processing for engine characteristic estimation by the engine characteristic estimation device.

Fig. 7 is a schematic diagram showing an example of engine characteristic estimation by the engine characteristic estimation device.

Description of the reference numerals

100: an engine characteristic estimating device; 110: a simulator; 120: an observer; 130: an engine characteristic estimating unit; 131: a difference calculation unit; 140: a calculation model updating unit; 150: a calculation model update mode execution unit; 200: an engine; 210: an engine main body; 220: an intake passage; 230: an exhaust passage; 240: a supercharger; x1: a first engine state parameter; x2: a second engine state parameter.

Detailed Description

Fig. 1 is a schematic diagram showing the overall configuration of an engine characteristic estimation device 100 according to the present embodiment. First, the engine 200, which is the target of the characteristic estimation by the engine characteristic estimation device 100, is driven by receiving the supply of fuel of an amount specified by the fuel supply amount U per one combustion, and generates power. The state parameter X0, which is a state variable of the engine 200, changes according to the operating state of the engine 200. Here, X0 is a vector including a plurality of parameters (the natural number n of 2 or more is the number of parameters, and each parameter is represented by X01, X02, …, and X0 n). What is selected as the state parameter X0 of the engine 200 can be appropriately determined according to the control target and the specification of the engine system, but typically, it is preferable to select the rotation speed, the exhaust pressure, the exhaust temperature, the exhaust amount, and the like of the engine main body. When a supercharger that increases the pressure of the air flowing into the engine body is used, it is preferable that the state parameter X0 include parameters of the supercharger, such as the number of revolutions, the intake air temperature, the intake air pressure, the intake air amount, the scavenging temperature, the scavenging pressure, and the scavenging amount.

In the engine characteristic estimation of the present embodiment, since it is more important to focus on the state parameter X0 inside the engine 200 than on the power generated outside the engine 200, the engine 200 is illustrated as a block that outputs the state parameter X0 based on the input of the fuel supply amount U per combustion.

The engine characteristic estimation device 100 is a device that estimates the characteristic of the engine 200, and includes a simulator 110 as a first calculation unit, an observer 120 as a second calculation unit, an engine characteristic estimation unit 130, a calculation model update unit 140, and a calculation model update mode execution unit 150.

The simulator 110 calculates a first engine state parameter X1 as a state variable of the engine 200 based on a first calculation model representing the characteristics of the engine 200 and a fuel supply amount U per one combustion supplied to the engine 200 at an arbitrary time point. Here, X1 is a vector including a plurality of parameters (the number of parameters is n, and each parameter is X11, X12, …, and X1n, similarly to the state parameter X0).

The observer 120 calculates a second engine state parameter X2 as a state variable of the engine 200 based on a second calculation model different from the first calculation model representing the characteristic of the engine 200, the motion data X0' relating to the motion of the engine 200 driven at an arbitrary point in time, and the fuel supply amount U per one combustion. Here, X2 is a vector including a plurality of parameters (the number of parameters is n, and elements of the vector are X21, X22, …, and X2n, similarly to the state parameter X0).

The motion data X0' as an input to the observer 120 is a part (1 or more and less than n) of the state parameters X0 included in the engine 200. In the actual engine 200, since it is not practical to provide sensors for measuring all of the n parameters included in the state parameter X0 due to the cost and the limitation in the installation, only a part of the parameter X0' is measured and used as an input to the observer 120. The basic function of the observer 120 is to complement this part of the parameters X0' and to estimate all the state parameters X2.

The engine characteristic estimating section 130 estimates the characteristic of the engine 200 at an arbitrary point in time based on the first engine state parameter X1 and the second engine state parameter X2. Specifically, the difference between the first engine state parameter X1 and the second engine state parameter X2 is calculated, and the characteristics of the engine 200 are estimated based on the difference.

The calculation model updating section 140 updates the second calculation model of the observer 120 based on the characteristic of the engine 200 estimated by the engine characteristic estimating section 130.

The calculation model update mode execution unit 150 executes an update mode for updating the second calculation model by the calculation model update unit 140.

Next, the above-described respective configurations will be described in more detail with reference to other drawings.

Fig. 2 is a schematic diagram showing the structure of the engine 200. In the present embodiment, the application of the engine 200 is not limited, and various engines 200 for ships, vehicles, aircrafts, and the like can be used. The engine 200 includes: an engine main body 210 that generates power by combustion of fuel; an intake passage 220 that supplies air for combusting fuel to the engine main body 210; an exhaust passage 230 for discharging burned gas in the engine main body 210; and a supercharger 240 for increasing the pressure of air supplied into the engine main body 210 through the intake passage 220.

The engine main body 210 includes: a combustion chamber 211 for combustion of fuel caused by air to occur; a fuel supply nozzle 212 for supplying fuel of an amount specified by a fuel supply amount U per one combustion into the combustion chamber 211; an intake valve 213 for controlling air supply from the intake passage 220 to the combustion chamber 211; an exhaust valve 214 for controlling air discharge from the combustion chamber 211 to the exhaust passage 230; a piston 215 linearly driven according to combustion of fuel in the combustion chamber 211; a crankshaft 216 that is rotationally driven in accordance with the linear motion of the piston 215; and a connecting rod 217 having one end fixed to the piston 215 and the other end fixed to the crankshaft 216, for converting the linear motion of the piston 215 into the rotational motion of the crankshaft 216. Further, although the fuel is directly supplied into the combustion chamber 211 through the fuel supply nozzle 212 as described above, in the case of using a highly volatile fuel such as gasoline, the fuel may be injected into the intake passage 220 and supplied into the combustion chamber 211 in a state of being mixed with air.

In the above configuration, engine 200 is driven by the following cycle. Here, the engine 200 is in an operating state by driving before the previous cycle, and the piston 215 repeatedly ascends and descends in accordance with the operation of the crankshaft 216 that continues to rotate by inertia.

(1) Air intake: the intake valve 213 opens, the exhaust valve 214 closes, and the piston 215 moves down, thereby supplying air from the intake passage 220 into the combustion chamber 211.

(2) Compression: the intake valve 213 is closed and the piston 215 is raised, whereby the air in the combustion chamber 211 is compressed.

(3) And (3) combustion: fuel of an amount specified by the fuel supply amount U per one combustion is supplied from the fuel supply nozzle 212 into the combustion chamber 211, and is combusted together with the compressed air. Thereby generating power and lowering the piston 215.

(4) Exhausting: exhaust valve 214 opens and piston 215 moves upward, thereby discharging combusted gas from combustion chamber 211 to exhaust passage 230.

The supercharger 240 is a so-called turbocharger, and includes a compressor 241 provided in the intake passage 220, a turbine 242 provided in the exhaust passage 230, and a shaft 243 coaxially coupling the compressor 241 and the turbine 242. The turbine 242 rotates by the gas discharged from the exhaust passage 230, and the rotation is transmitted to the compressor 241 via a shaft 243. Since the air supplied to intake passage 220 is compressed by compressor 241 rotating in this manner, the pressure of the air supplied to combustion chamber 211 can be increased.

In fig. 2, a so-called four-stroke engine in which four strokes of (1) intake, (2) compression, (3) combustion, and (4) exhaust are taken as a primary cycle is exemplified, but the type of engine is not limited to this, and various types of engines can be used in the present embodiment. For example, a so-called two-stroke engine shown in fig. 3 can be used (the same reference numerals are given to the components corresponding to fig. 2).

As in the case of the four-stroke engine, the engine body 210 of the two-stroke engine linearly drives the piston 215 by combustion of the fuel in the combustion chamber 211 and converts the driving into rotational power of the crankshaft 216. The main structure of the two types of engines is almost the same, but one difference of the two-stroke engine is that a scavenging passage 219 is provided in the engine body 210, and the scavenging passage 219 connects a crankcase 218 that houses a crankshaft 216 and a combustion chamber 211.

In the illustrated state in which the piston 215 is lowered, gas can flow through a path passing through the crankcase 218, the scavenging passage 219, the combustion chamber 211, and the exhaust passage 230, and fresh air in the crankcase 218 flows into the combustion chamber 211 through the scavenging passage 219, and the combusted gas is discharged (scavenged) to the exhaust passage 230 by the momentum thereof.

Next, if the piston 215 rises, the scavenging passage 219 and the exhaust passage 230 are closed, the combustion chamber 211 is sealed, and the pressure thereof rises. Then, fuel is supplied from the fuel supply nozzle 212 into the combustion chamber 211 having a high pressure to cause combustion, and power for lowering the piston 215 again is generated. On the other hand, when the piston 215 moves up, the crankcase 218 communicates with the intake passage 220, and fresh air flows into the crankcase 218 from the intake passage 220. Thus, when the piston 215 moves up, combustion in the combustion chamber 211 and intake air in the crankcase 218 are simultaneously performed.

As described above, in the two-stroke engine, one cycle is completed with two strokes in total of one descent and one ascent of the piston 215. In such a two-stroke engine, if the supercharger 240 shown in fig. 2 is used, the intake pressure applied to the crankcase 218 when the piston 215 is raised and the scavenging pressure applied to the combustion chamber 211 when the piston 215 is lowered can be increased.

As the two-stroke engine, an engine having a scavenging receiver for receiving air for scavenging as disclosed in patent document 2 may be used. In this case, similarly to the above description of scavenging in fig. 3, in a state where the piston (41: the reference numeral (the same shall apply hereinafter) in patent document 2) is lowered, the gas can flow through a path passing through the scavenging receiver (2), the scavenging port (17) corresponding to the crankcase 218 and the scavenging passage 219, the cylinder (1) corresponding to the combustion chamber 211, and the exhaust duct (6) corresponding to the exhaust passage 230, and the scavenging operation of discharging the burned gas to the exhaust duct is performed by the momentum of the fresh air in the scavenging receiver flowing into the cylinder through the scavenging port. In addition, if the supercharger 240 is used in this structure, the pressure of scavenging in the scavenging receiver can be increased.

As described above, the present embodiment can be applied to various types of engines 200, but is particularly preferably applied to a marine engine having a rated rotational speed of 1000 rpm or less. In general, a marine engine can be driven at a lower rated rotation speed than a vehicle engine. In particular, in a large ship, it takes time until the power generated by the engine is reflected in the actual operation of the ship, and therefore, accurate engine driving is required. As described above, in the marine engine, it is highly required to accurately estimate the characteristic change and state of the engine and to perform accurate driving, and the engine characteristic estimation device 100 of the present embodiment is preferably used.

Here, the following characteristics are exemplified as the characteristics of the engine 200 estimated by the engine characteristic estimation device 100.

Combustion efficiency: combustion efficiency in the combustion chamber 211. Also known as thermal efficiency.

Power transmission efficiency: the ratio of the effective torque obtained by subtracting the loss of each mechanical part to the torque generated by the engine main body 210. Also known as mechanical transmission efficiency.

Dynamic characteristics: the temporal relationship between the parameters is taken into account. Responsiveness of pressure to temperature changes, etc.

Efficiency of the supercharger 240: the efficiency of the compressor 241, the efficiency of the turbine 242, etc.

Interference effects: the temperature (atmospheric temperature) and pressure (atmospheric pressure) of the outside air taken in by the engine 200, and a load caused by waves and the like flowing in from a propeller as a driving object in the marine engine.

The disturbance is an important disturbance that has a large influence on the actual operation of engine 200, and can be treated in the same way as other characteristics in a mathematical model described later in relation to fig. 4.

In the engine 200 described above by way of example, the state parameter X0 used for characteristic estimation and control can be constituted by the following parameters, for example.

Parameters related to the operation of the engine main body 210:

rotational speed of crankshaft 216 (rotational speed Ne of engine body 210)

Temperature of gas discharged from exhaust passage 230 (exhaust temperature Tex of engine body 210)

Pressure of gas discharged from the exhaust passage 230 (exhaust pressure Pex of the engine body 210)

The flow rate of gas discharged from the exhaust passage 230 (the exhaust gas amount Gex of the engine body 210)

Parameters related to the operation of the supercharger 240:

rotational speeds of compressor 241, turbine 242, and shaft 243 (rotational speed Ntc of supercharger 240)

In the four-stroke engine of fig. 2 or the like in which the scavenging operation is not performed, the temperature of the air supplied from the intake passage 220 to the combustion chamber 211 via the supercharger 240 (the intake air temperature Tb of the supercharger 240)

In the four-stroke engine of fig. 2 or the like in which the scavenging operation is not performed, the pressure of the air supplied from the intake passage 220 to the combustion chamber 211 via the supercharger 240 (the supply pressure Pb of the supercharger 240)

In the four-stroke engine of fig. 2 or the like in which the scavenging operation is not performed, the flow rate of air supplied from the intake passage 220 to the combustion chamber 211 via the supercharger 240 (the air supply amount Gb of the supercharger 240)

In the two-stroke engine of fig. 3, patent document 2, or the like that performs the scavenging operation, the pressure of the air supplied from the scavenging passage 219 to the combustion chamber 211 and the temperature of the air in the scavenging receiver (the scavenging temperature Ts of the supercharger 240)

In the two-stroke engine of fig. 3, patent document 2, or the like that performs the scavenging operation, the pressure of the air supplied from the scavenging passage 219 to the combustion chamber 211 and the pressure of the air in the scavenging receiver (the scavenging pressure Ps of the supercharger 240)

In the two-stroke engine of fig. 3, patent document 2, or the like that performs the scavenging operation, the pressure of the air supplied from the scavenging passage 219 to the combustion chamber 211 and the flow rate of the air in the scavenging receiver (the scavenging amount Gs of the supercharger 240)

In the case where the supercharger 240 is not provided, since the intake air to the combustion chamber 211 (in the case of a four-stroke engine) and the scavenging to the combustion chamber 211 (in the case of a two-stroke engine) are operations of the engine main body 210, the intake air temperature Tb, the intake air pressure Pb, the intake air amount Gb, the scavenging temperature Ts, the scavenging pressure Ps, and the scavenging amount Gs are parameters relating to the operations of the engine main body 210.

Although each of the above parameters can be measured by providing an appropriate sensor, in the actual engine 200, it is not practical to measure all the parameters due to the cost and the limitation in installation, and a part of the parameter X0' is measured and used as an input to the observer 120. The selection of the parameter X0' to be measured can be appropriately determined according to the control target or specification of the engine system, but is preferably selected on the basis of the following criteria, for example.

At least one parameter is measured from each of the parameters (engine body data) related to the operation of the engine body 210 and the parameters (supercharger data) related to the operation of the supercharger 240. The engine body data include the rotation speed Ne, the exhaust temperature Tex, the exhaust pressure Pex, and the exhaust gas amount Gex, which are listed above. The supercharger data may be exemplified by the previously mentioned rotation speed Ntc, intake air temperature Tb (in the case of a four-stroke engine), intake air pressure Pb (in the case of a four-stroke engine), intake air amount Gb (in the case of a four-stroke engine), scavenging temperature Ts (in the case of a two-stroke engine), scavenging pressure Ps (in the case of a two-stroke engine), and scavenging amount Gs (in the case of a two-stroke engine). If the parameters thus measured are selected, the observer 120 can estimate the state of the entire system of the engine 200 with high accuracy based on the measurement data of each of the engine body 210 and the supercharger 240.

As a further criterion, at least one parameter is determined from a parameter relating to the mechanical action of the engine 200 (mechanical data) and a parameter relating to the thermodynamic state of the engine 200 (thermodynamic data), respectively. The mechanical data may be the rotation speed Ne of the engine body 210 and the rotation speed Ntc of the supercharger 240, which are listed above. The thermodynamic data include the exhaust temperature Tex, the exhaust pressure Pex, the exhaust gas amount Gex, the intake air temperature Tb (in the case of a four-stroke engine), the intake air pressure Pb (in the case of a four-stroke engine), the intake air amount Gb (in the case of a four-stroke engine), the scavenging temperature Ts (in the case of a two-stroke engine), the scavenging pressure Ps (in the case of a two-stroke engine), and the scavenging amount Gs (in the case of a two-stroke engine), which have been described above. If the parameters thus measured are selected, the observer 120 can estimate the state with high accuracy in a manner that takes into account the mechanical and thermodynamic aspects of the engine 200, based on the respective measured data of the mechanical and thermodynamic data.

In practical designs, it is desirable to select measurement parameters that satisfy both of the above criteria. For example, the rotation speed Ne of the engine body 210 and the scavenging pressure Ps of the supercharger 240 may be selected as the measurement parameters. Here, the rotation speed Ne is engine body data and mechanical data, and the scavenging pressure Ps is supercharger data and thermodynamic data, and both the above-described two criteria are satisfied.

The fuel supply amount U per one combustion, which is the drive input to the engine 200, is set based on the measurement data of the rotation speed Ne of the engine main body 210. That is, assuming that Ne0 is the target rotation speed of the engine main body 210, the difference between Ne as the measurement value and Ne0 as the target value is calculated, and the fuel supply amount U per combustion is set such that the difference is reduced, based on a predetermined table or algorithm.

Next, the structures of the simulator 110 and the observer 120 are explained with reference to fig. 4.

The simulator 110 receives the fuel supply amount U per combustion supplied to the engine 200, and calculates a first engine state parameter X1 (X11, X12, …, X1n) which is a vector having n elements, based on a first calculation model. Here, the number of elements of the first engine state parameter X1 and the state parameter X0 of the engine 200 is equal, and each element of X1 is an estimated value of each element of X0. For example, when the first element X01 of the state parameter X0 of the engine 200 is the rotation speed Ne and the second element X02 is the scavenging pressure Ps, the first element X11 of the first engine state parameter X1 is an estimated value of the rotation speed Ne and the second element X12 is an estimated value of the scavenging pressure Ps.

The first calculation model of the simulator 110 represents the engine characteristics of the engine 200 at a predetermined reference time point, and is a model simulating the state parameter X0 of the engine 200 at the reference time point. Here, since the first calculation model is set on the assumption that there is an ideal state in which there is no disturbance to the engine 200, the first engine state parameter X1 as the calculation result indicates the ideal state parameter X0 of the engine 200 at the reference time point. As a reference time point for specifying the first calculation model, an initial time point of the engine 200 is typically selected. In this case, the first engine state parameter X1 is a result of simulation of the ideal state parameter X0 in the case where the ideal engine 200 with no characteristic degradation from the initial state is placed in a non-interference state.

Next, the state space expression of the engine 200 and the observer 120 having the state space expression corresponding thereto are explained. In fig. 4, it is assumed that the engine 200 is a linear system defined by the system coefficient matrix a. Specifically, the state of the engine 200 is described by the following equation.

dX0/dt＝A·X0+B·U

X0'＝C·X0

Here, each parameter is as follows.

U: fuel supply amount to engine 200

X0: state parameters of engine 200

X0': input to observer 120 (measured for some of the operating parameters included in state parameter X0 as described above)

A: system coefficient matrix of engine 200

B: inputting an input vector of U into a system

C: extracting an output vector of X0' from X0

Further, the illustrated state space expression of the engine 200 shows the above equation.

The observer 120 has the same state space expression as the engine 200 described above. That is, the observer 120 has the same system coefficient matrix a, input vector B, and output vector C as those of the engine 200. A large difference from the engine 200 is that the motion data X0' of the engine 200 is input to the observer 120 via the observer gain H. In this way, the observer 120 has a second calculation model defined by the system coefficient matrix a and the observer gain H, and calculates the second engine state parameter X2 with the fuel supply amount U and the motion data X0' per combustion as inputs.

The second engine state parameter X2 is a vector having n elements (X21, X22, …, X2 n). Like the first engine state parameter X1, the number of elements of the second engine state parameter X2 is equal to the number of elements of the state parameter X0 of the engine 200, and each element of X2 is an estimated value of each element of X0. For example, when the first element X01 of the state parameter X0 of the engine 200 is the rotation speed Ne and the second element X02 is the scavenging pressure Ps, the first element X21 of the second engine state parameter X2 is an estimated value of the rotation speed Ne and the second element X22 is an estimated value of the scavenging pressure Ps.

Unlike the simulator 110 that calculates the ideal state parameter X1 at a certain reference time point, the observer 120 calculates the real-time state parameter X2 of the engine 200. In the case where all the state parameters X0 of the engine 200 are measured, the observer 120 is not necessarily provided, but it is impractical to measure all the parameters, and therefore the observer 120 is used to estimate the state parameters X2.

Although the above description has been given by taking the case where the engine 200 is a linear system as an example, the observer 120 can be configured similarly in the case of a nonlinear system. That is, the observer 120 may include an element (a system coefficient matrix in the example of fig. 4) representing the system characteristic of the engine 200 in the second calculation model, and whether the engine 200 is a linear system or a nonlinear system is not essential in the present embodiment.

Fig. 5 is a schematic diagram showing the structure of the engine characteristic estimating section 130. The engine characteristic estimating unit 130 includes a difference calculating unit 131, an absolute value calculating unit 132, a weighting calculating unit 133, an adding unit 134, and a threshold comparing unit 135, and sets the difference calculation result of the difference calculating unit 131 as the characteristic estimation output of the engine 200.

The difference calculation unit 131 calculates the difference between the first engine state parameter X1 and the second engine state parameter X2. Specifically, differentiators 131-1, 131-2, … and 131-n equal in number to the number n of elements X1 and X2 are provided. Each differentiator calculates differences e1, e2, … and en between the parameters corresponding to X1 and X2. As in the above example, when the first element X01 of the state parameter X0 is the rotation speed Ne and the second element X02 is the scavenging pressure Ps, the first differentiator 131-1 calculates the difference e1 between X11, which is the estimated value of the rotation speed Ne, and X21, which is the estimated value of the rotation speed Ne, and the second differentiator 131-2 calculates the difference e2 between X12, which is the estimated value of the scavenging pressure Ps, and X22, which is the estimated value of the scavenging pressure Ps. These difference calculation results e1, e2, …, en are supplied to the calculation model updating unit 140 at the subsequent stage as characteristic estimation outputs of the engine 200.

The absolute value calculation unit 132, the weighting calculation unit 133, the addition unit 134, and the threshold comparison unit 135 trigger the calculation model update process of the calculation model update unit 140 at the subsequent stage according to the series of calculation results.

The absolute value calculation unit 132 calculates the absolute values of the differences e1, e2, …, and en from the difference calculation unit 131.

The weighting calculation unit 133 multiplies the n calculation results from the absolute value calculation unit 132 by predetermined weights w1, w2, …, wn. Here, each weight is appropriately set according to the importance of each parameter in the engine characteristic estimation of the present embodiment. For example, in the above example, when the rotation speed Ne as the first parameter is emphasized more than the scavenging pressure Ps as the second parameter in the engine characteristic estimation, the first weight w1 is preferably set to be greater than the second weight.

The adder 134 adds the n calculation results from the weighting calculator 133.

The threshold value comparing unit 135 compares the operation result from the adding unit 134 with a predetermined threshold value. When the calculation result from the adder 134 exceeds the threshold, a trigger signal T for the calculation model update process of the calculation model update unit 140 of the subsequent stage is generated.

As described above, the trigger signal T for the computation model update unit 140 is generated only when the differences e1, e2, …, en from the difference calculation unit 131 satisfy the predetermined reference, and therefore, the frequency of computation model update processing can be prevented from becoming too high, and the stability of the system can be improved. The threshold comparing unit 135 may be configured to generate the trigger signal T when a state in which the calculation result from the adding unit 134 exceeds the threshold continues for a predetermined time. In this case, the trigger signal T is not generated due to an instantaneous abnormal value, and therefore, the stability of the system is further improved.

Note that the series of configurations of the absolute value calculation unit 132, the weighting calculation unit 133, the addition unit 134, and the threshold comparison unit 135 described above is merely an example, and various configurations can be adopted as long as the trigger signal T for the calculation model update unit 140 is generated based on the difference calculation results e1, e2, …, and en of the difference calculation unit 131. For example, when only one of the plurality of difference operation results (e.g., e2) is important, the trigger signal T may be generated based on only the magnitude of the difference operation result without considering the other difference operation results. In this case, the weighting unit 133 and the adding unit 134 are not required.

Returning to fig. 1, the calculation model updating unit 140 receives the trigger signal T from the engine characteristic estimating unit 130, receives the characteristic estimation outputs e1, e2, …, and en from the engine characteristic estimating unit 130 as inputs, and updates the second calculation model of the observer 120 based on a predetermined algorithm. As already explained in fig. 4, the second calculation model is a calculation model including a system coefficient matrix a and an observer gain H. Here, the calculation model updating unit 140 updates at least one of the system coefficient matrix a and the observer gain H based on a predetermined algorithm. The system coefficient matrix a as a matrix and the observer gain H as a vector have a plurality of elements, respectively, and therefore at least one of the elements is updated.

In particular, when coping with a characteristic change of engine 200 due to an external environment change such as aged deterioration or an intake air temperature, it is preferable to update system coefficient matrix a. When the characteristics of the engine 200 change, that is, when the combustion efficiency, the power transmission efficiency, the dynamic characteristics, the supercharger efficiency, the disturbance influence, and the like change as described above, the system coefficient matrix a of the engine 200 in fig. 4 changes, and therefore the system coefficient matrix a of the observer 120 may be updated in accordance with the change. Thus, the second calculation model of the observer 120 reflects the characteristics of the engine 200, and therefore the state estimation accuracy when the state of the engine 200 is estimated by the observer 120 can be improved.

Further, the algorithm used in the calculation model update process by the calculation model update unit 140 can be designed appropriately with the goal of making the second calculation model of the observer 120 better reflect the actual characteristics of the engine 200. As a simple example, the following error-trial type algorithm can be cited. In this algorithm, a plurality of update processing options prepared in advance are tried in sequence, and the options that improve the characteristic estimation outputs e1, e2, …, en most are adopted. The algorithm is not limited to a pre-programmed algorithm, and may be updated by machine learning according to an actual processing result.

The calculation model update mode execution unit 150 executes an update mode for updating the second calculation model by the calculation model update unit 140. While the update mode is not being executed, the calculation model update unit 140 does not update the second calculation model even if the trigger signal T is generated from the engine characteristic estimation unit 130. On the other hand, if the trigger signal T is generated from the engine characteristic estimating section 130 while the update mode is being executed, the calculation model updating section 140 performs the update of the second calculation model based on the characteristic estimation outputs e1, e2, …, en. Further, the update mode can be executed not only based on a user operation but also automatically at a predetermined frequency. By providing such a dedicated update mode, it is possible to prevent the frequency of the calculation model update processing from becoming excessively high, thereby improving the stability of the system. On the other hand, it is needless to say that the following configuration is possible: the update process of the second calculation model is performed at any time based on the trigger signal T from the engine characteristic estimation unit 130 without providing such a dedicated update mode.

Fig. 6 shows a process flow of engine characteristic estimation by the engine characteristic estimation device 100 having the above structure.

In step S10, it is determined whether or not the update mode is executed by the calculation model update mode execution unit 150.

In the case where the update mode is executed, the simulator 110 calculates the first engine state parameter X1 in step S21, and the observer 120 calculates the second engine state parameter X2 in parallel with this in step S22.

In step S30, the engine characteristic estimating unit 130 calculates differences e1, e2, …, and en (characteristic estimation outputs) between the elements of the first engine state parameter X1 and the second engine state parameter X2.

In step S40, the engine characteristic estimating section 130 determines whether or not the trigger signal T for the calculation model updating section 140 is generated based on the characteristic estimation outputs e1, e2, …, en.

When the trigger signal T is generated, the calculation model updating unit 140 updates the second calculation model of the observer 120 based on the characteristic estimation outputs e1, e2, …, and en in step S50.

Fig. 7 is a schematic diagram showing an example of the engine characteristic estimation performed by the engine characteristic estimation device 100 described above. As in fig. 1, the engine 200, the simulator 110, and the observer 120 are represented as modules that output state parameters of the engine 200. The following description is made on the assumption that an ideal state is not affected by disturbance on engine 200, but the present embodiment can be applied to a case where disturbance is present as well.

Fig. 7 (a) shows an initial state of the engine 200.

The engine 200 is described by a system coefficient matrix a0 showing the characteristics of its initial state, and outputs a state parameter X0 for input of a fuel supply amount U per combustion.

The simulator 110 calculates a first engine state parameter from the input of the fuel supply amount U per one combustion based on a first calculation model representing the characteristics of the engine 200 in the initial state. The simulator 110 correctly reproduces the engine 200 in the initial state, and the first engine state parameter as its output coincides with the state parameter X0 of the engine 200 in the initial state described above.

The observer 120 calculates a second engine state parameter from the input of the fuel supply amount U per combustion and the motion data X0' based on a second calculation model including the same system coefficient matrix a0 as the engine 200 in the initial state. Here, the observer 120 estimates the initial state of the engine 200 based on the system coefficient matrix a0, and if it is assumed that the estimation accuracy thereof is 100%, the second engine state parameter coincides with the state parameter X0 of the engine 200 in the above-described initial state.

Thus, in the initial state of fig. 7 (a), the outputs of the engine 200, the simulator 110, and the observer 120 coincide. Then, the output of the difference calculation unit 131 that calculates the difference between the outputs of the simulator 110 and the observer 120 is zero.

Fig. 7 (B) shows a state in which the characteristics of engine 200 have changed due to degradation with time from the initial state.

Due to the deterioration of the engine 200, the system coefficient matrix of the engine 200 changes from a0 in the initial state to a 1. In addition, due to the change of the system coefficient matrix, the state parameter thereof also changes from X0 of the initial state to X.

The first computational model of the simulator 110 has not changed from the initial state. Since the input to the simulator 110, i.e., the fuel supply amount U per combustion is a drive input amount that is not related to the deterioration of the engine 200, the first engine state parameter, which is the output of the simulator 110, is represented by X0 without being changed from the initial state.

The second computational model of the observer 120 is also unchanged from the initial state, including the system coefficient matrix a0 for the initial state of the engine 200. On the other hand, since the motion data X 'input from the engine 200 to the observer 120 is affected by the deterioration of the engine 200, it changes from the input X0' to the observer 120 in the initial state. As a result, the second engine state parameter, which is the output of the observer 120, changes to X2, which is different from X0 in the initial state.

In this way, in the degraded state shown in fig. 7B, the first engine state parameter X1 (equal to X0) deviates from the second engine state parameter X2 (different from X0), and the output of the difference arithmetic unit 131 is not zero (X2-X0). As described above, according to the engine characteristic estimation device 100 of the present embodiment, the deterioration of the engine 200, that is, the change from a0 to a1 of the system coefficient matrix can be estimated based on the deviation.

Fig. 7 (C) shows a state in which the calculation model updating unit 140 performs the update process of the observer 120 on the basis of the deterioration of the engine 200 estimated in fig. 7 (B). The calculation model updating unit 140 estimates the system coefficient matrix a1 after the engine 200 is deteriorated based on the deviation data X2-X0 calculated in the deteriorated state, and updates the system coefficient matrix in the second calculation model of the observer 120 from a0 to a 1. Thereby, the system coefficient matrix of the observer 120 reflects the characteristic of the degraded engine 200, and the estimation accuracy of the observer 120 on the engine 200 improves. Here, if the estimation accuracy is assumed to be 100%, the output of the observer 120 completely coincides with the output X of the engine 200.

As described above, according to the engine characteristic estimation device 100 of the present embodiment, the deterioration of the engine 200 is estimated, and the update process of the observer 120 is performed based on the estimated deterioration, whereby the state estimation accuracy of the engine 200 can be improved.

Further, in the update state of fig. 7 (C), although a stable deviation X-X0 is generated between the output X of the observer 120 and the output X0 of the simulator 110, there is no problem. After the update state, if the deviation remains at X-X0, it is known that the engine 200 is not further deteriorated, and therefore, the calculation model updating unit 140 does not need to update the calculation model any more, and the system can be normally operated in this state. On the other hand, when the engine 200 is further deteriorated and the deviation from X-X0 is further changed, the calculation model updating unit 140 updates the calculation model further, so that the latest characteristics of the engine 200 can be reflected in the system coefficient matrix of the observer 120.

In fig. 7 (a), the case where the output of the simulator 110 and the output of the observer 120 match in the initial state is described as an example, but the output of the simulator 110 and the output of the observer 120 may be different. Here, if the initial output of the simulator 110 is set to X0 and the initial output of the observer 120 is set to X, the deviation of the initial state is represented as X-X0. This is substantially the same as the case where there is a stable deviation X-X0 as illustrated in fig. 7 (C).

In addition, in order to eliminate the stable deviation in appearance as described above, the first calculation model of the simulator 110 can also be updated. That is, when there is a steady deviation X2-X1 between the first engine state parameter X1 of the simulator 110 and the second engine state parameter X2 of the observer 120 at a predetermined reference time point such as the initial time point of fig. 7 (a) or the update time point of fig. 7 (C), the first calculation model of the simulator 110 is updated, and the output of the simulator 110 is corrected from X1 to X2. Accordingly, since the output of the simulator 110 and the output of the observer 120 are both X2, the above-described steady deviation can be visually eliminated.

Even when the first calculation model of the simulator 110 is updated in this manner, the first calculation model corresponding to the initial state is preferably continuously maintained. The initial state is an absolute reference state when the characteristic of engine 200 is estimated or controlled, and therefore it is preferable to make the initial state always referable. For example, when the first calculation model of the simulator 110 is updated in the update state shown in fig. 7 (C), the first calculation model corresponding to the initial state can be directly held, and a new first calculation model corresponding to the update state can be added. In this case, the simulator 110 calculates a first engine state parameter corresponding to the initial state and a first engine state parameter corresponding to the updated state, respectively. Then, the difference calculation unit 131 calculates the difference between each of the first engine state parameters and the second engine state parameters, thereby estimating the deterioration of the engine 200 from the initial state and the deterioration from the updated state.

In addition to the above description, the present embodiment has, for example, the following operations and effects.

The fuel supply amount U per combustion used in the calculation by the simulator 110 is data that is not affected by the change in the characteristics of the engine 200 due to the aged deterioration, the change in the external environment such as the intake air temperature, and the like, and therefore the first engine state parameter X1 that is the calculation result thereof is not affected by the change in the characteristics of the engine 200. In contrast, since the operation data of the engine 200 used for the calculation of the observer 120 is data influenced by the characteristic change of the engine 200, the second engine state parameter X2 as the calculation result is influenced by the characteristic change of the engine 200. In this way, by using the characteristic estimation output, which is the difference between the two engine state parameters X1 and X2 having different influences of the engine characteristic change, the engine characteristic estimation unit 130 can accurately estimate the characteristic change of the engine 200 (the change from a0 to a1 in the system coefficient matrix). Further, even when the characteristics of engine 200 are not changed, the state of engine 200 can be estimated with high accuracy by using two calculation units having different calculation models at the same time.

By using a model representing the characteristics of the initial state of the engine 200 as the first calculation model of the simulator 110 and the second calculation model of the observer 120, the characteristic change of the engine 200 from the initial state can be effectively estimated.

By providing the calculation model updating unit 140 that updates the second calculation model of the observer 120 based on the characteristic of the engine 200 estimated by the engine characteristic estimating unit 130, the accuracy in estimating the state of the engine 200 by the observer 120 can be maintained high even when the engine 200 is deteriorated.

Although there are parameters that represent nonlinearity in the parameters that constitute the first engine state parameter X1 and the second engine state parameter X2, there are cases where nonlinearity can be reduced by calculating the difference between these parameters. In the present embodiment, the difference calculation unit 131 can perform characteristic estimation of the engine 200 by calculating the difference between the first engine state parameter X1 and the second engine state parameter X2 to convert the difference into an easy-to-handle form in which nonlinearity is reduced.

The present invention has been described above based on the embodiments. As can be appreciated by those skilled in the art: the embodiments are illustrative, and various modifications can be made to the combination of these respective constituent elements and the respective processing procedures, and such modifications are also within the scope of the present invention.

In the present embodiment, the engine characteristic estimating unit 130 includes the difference calculating unit 131 that calculates the difference between the first engine state parameter X1 and the second engine state parameter X2, and estimates the characteristic of the engine 200 based on the difference, but may estimate the characteristic of the engine 200 based on other calculations. For example, since the simulator 110 and the observer 120 have different input data and calculation models, there is a possibility that the estimation accuracy cannot be maximized even when a simple difference between X1 and X2 is used. In this case, the function for performing the calculation of the characteristic estimation in the engine characteristic estimation unit 130 can be optimized according to the difference between the input data of the both and the calculation model. If such a function is generalized and is f (X1, X2), the engine characteristic estimating unit 130 can be understood as a computing unit that inputs X1 and X2 and outputs the characteristic estimation output f (X1, X2). In the example of the difference operation described in the embodiment, f (X1, X2) is X2 to X1.

In the embodiment, the engine characteristic estimation device 100 including the simulator 110 as the first calculation unit that calculates the first engine state parameter X1 using the first calculation model, the observer 120 as the second calculation unit that calculates the second engine state parameter X2 using the second calculation model, and the engine characteristic estimation unit 130 that estimates the characteristic of the engine 200 based on X1 and X2 has been described, but the following engine state estimation device may be configured: the engine characteristic estimating unit 130 is not provided, and the state of the engine 200 is estimated by the simulator 110 and the observer 120. According to this engine state estimation device, the state of the engine 200 can be estimated with high accuracy by using two calculation units having different calculation models at the same time. For example, in the case where the scavenging pressure is estimated as the state parameter of the engine 200, since the estimated values thereof are included in both the first engine state parameter X1 from the simulator 110 and the second engine state parameter X2 from the observer 120, the scavenging pressure can be estimated with high accuracy by calculating the average value or the like of the two estimated values. Generally, an engine state estimating unit may be configured as a calculation unit that prepares a function g (X1, X2) for state estimation that is most suitable for engine 200 according to the difference between the calculation models, and outputs a state estimation output g (X1, X2) with X1 and X2 as inputs. In an example of calculating the average value, g (X1, X2) is (X1+ X2)/2. Further, the engine state estimating unit may be configured to: the operation based on g (X1, X2) is performed so as to include not only the transient data of X1 and X2 but also the history data of X1 and X2 in the past fixed period.

The functional configuration of each device described in the embodiments can be realized by hardware resources, software resources, or cooperation of hardware resources and software resources. As the hardware resources, a processor, ROM, RAM, or other LSI can be used. As the software resource, a program such as an operating system or an application program can be used.

In the embodiments disclosed in the present specification, in which a plurality of functions are provided in a distributed manner, a part or all of the plurality of functions may be provided in a concentrated manner, and conversely, an embodiment in which a plurality of functions are provided in a concentrated manner may be provided in which a part or all of the plurality of functions are distributed. The functions may be integrated or distributed, and the functions may be configured to achieve the object of the invention.

Claims

1. An engine characteristic estimation device is provided with:

a first calculation unit that calculates a first engine state parameter as a state variable of an engine based on a first calculation model representing a characteristic of the engine and a fuel supply amount supplied to the engine at an arbitrary time point;

a second calculation unit that calculates a second engine state parameter that is a state variable of the engine, based on a second calculation model that is different from the first calculation model and that represents a characteristic of the engine, and on operation data relating to an operation of the engine that is driven at the arbitrary time point; and

an engine characteristic estimating section that estimates a characteristic of the engine at the arbitrary time point based on the first engine state parameter and the second engine state parameter.

2. The engine characteristic estimation device according to claim 1,

the first calculation model represents a characteristic of an initial state of the engine.

3. The engine characteristic estimation device according to claim 1 or 2,

the engine characteristic estimating device further includes a calculation model updating unit that updates the second calculation model based on the characteristic of the engine estimated by the engine characteristic estimating unit.

4. The engine characteristic estimation device according to claim 3,

the second calculation model includes a system coefficient matrix representing characteristics of the second calculation section,

the calculation model updating section updates at least one element in the system coefficient matrix.

5. The engine characteristic estimation device according to claim 3,

the engine characteristic estimating unit includes a difference calculating unit that calculates a difference between the first engine state parameter and the second engine state parameter, and estimates the characteristic of the engine based on the difference,

the calculation model update unit updates the second calculation model when the difference exceeds a predetermined threshold value.

6. The engine characteristic estimation device according to claim 5,

the calculation model update unit updates the second calculation model when a state in which the difference exceeds the threshold value continues for a predetermined time.

7. The engine characteristic estimation device according to claim 3,

further comprising a calculation model update mode execution unit that executes an update mode for updating the second calculation model by the calculation model update unit,

the calculation model updating portion updates the second calculation model based on the characteristic of the engine estimated by the engine characteristic estimating portion in the update mode.

8. The engine characteristic estimation device according to claim 1,

the engine characteristic estimating unit estimates at least one of combustion efficiency, power transmission efficiency, dynamic characteristics, efficiency of a supercharger for increasing pressure of air flowing into an engine main body, and influence of disturbance on the engine, as the characteristic of the engine.

9. The engine characteristic estimation device according to claim 1,

the engine characteristic estimating unit includes a difference calculating unit that calculates a difference between the first engine state parameter and the second engine state parameter, and estimates the characteristic of the engine based on the difference.

10. The engine characteristic estimation apparatus according to claim 9,

the difference calculation unit includes a plurality of differentiators,

the plurality of differentiators calculate differences between the plurality of parameters included in the first engine state parameter and the plurality of parameters included in the second engine state parameter corresponding thereto.

11. The engine characteristic estimation device according to claim 1,

the operation data includes engine body data related to an operation of an engine body and supercharger data related to an operation of a supercharger for increasing a pressure of air flowing into the engine body.

12. The engine characteristic estimation apparatus according to claim 11,

the engine body data includes at least one of a rotation speed, an exhaust temperature, an exhaust pressure, and an exhaust amount of the engine body.

13. The engine characteristic estimation device according to claim 11 or 12,

the supercharger data contains at least one of a rotation speed, an air supply temperature, an air supply pressure, an air supply amount, a scavenging temperature, a scavenging pressure, and a scavenging amount of the supercharger.

14. The engine characteristic estimation device according to claim 1,

the motion data includes mechanical data relating to mechanical motion of the engine and thermodynamic data relating to thermodynamic state of the engine.

15. The engine characteristic estimation apparatus according to claim 14,

the mechanical data includes at least one of a rotational speed of an engine main body and a rotational speed of a supercharger for increasing a pressure of air flowing into the engine main body.

16. The engine characteristic estimation apparatus according to claim 14 or 15,

the thermodynamic data includes at least one of exhaust temperature, exhaust pressure, exhaust volume, supply air temperature, supply air pressure, supply air volume, scavenging temperature, scavenging pressure, and scavenging volume.

17. The engine characteristic estimation device according to claim 1,

the second calculation portion calculates the second engine state parameter based on the fuel supply amount in addition to the operation data.

18. The engine characteristic estimation device according to claim 1,

the engine is a marine engine having a rated rotation speed of 1000 rpm or less.

19. An engine characteristic estimation method comprising the steps of:

a first calculation step of calculating a first engine state parameter as a state variable of an engine based on a first calculation model representing a characteristic of the engine and a fuel supply amount supplied to the engine at an arbitrary time point;

a second calculation step of calculating a second engine state parameter as a state variable of the engine based on a second calculation model different from the first calculation model representing a characteristic of the engine and action data on an action of the engine driven at the arbitrary time point; and

an engine characteristic estimating step of estimating a characteristic of the engine at the arbitrary time point based on the first engine state parameter and the second engine state parameter.

20. A computer-readable storage medium storing an engine characteristic estimation program for causing a computer to execute:

21. An engine state estimation device is provided with:

an engine state estimating unit that estimates a state of the engine at the arbitrary time point based on the first engine state parameter and the second engine state parameter.