JP2020186730A - Engine control method, engine control program, and engine control device using engine state observation device - Google Patents

Abstract

To provide an engine control method, an engine control program, and an engine control device that use an engine state observation device to enable the keeping of the engine performance at the maximum efficiency in the unsteady state by using an engine model to estimate the excess air rate and controlling the engine on the basis of the estimated excess air rate.SOLUTION: An engine state observation device 15 sets an initial state value for calculating an air amount Ga of an engine 1, and inputs the initial state value to an engine model of the engine state observation device linked with a propeller model that detects the rotation speed ne of the engine that drives the propeller of a ship as a sensor value and estimates the load in consideration of the disturbance in the actual sea area; determines a fuel amount Gf on the basis of the engine speed ne and load in the engine model; determines the air amount Ga to the engine 1 on the basis of observation parameters of the engine model obtained from the initial state value and the sensor value; estimates an excess air rate λ from the fuel amount Gf and the air amount Ga to the engine 1; and when the estimated excess air rate λ becomes lower than the value at the initial stage of operation, controls at least the exhaust valve opening EVC of an exhaust valve 6 so as to recover the excess air rate λ.SELECTED DRAWING: Figure 2

Description

The present invention relates to an engine control method, an engine control program, and an engine control device that control an engine provided with an exhaust valve and fuel adjusting means by using an engine state observer that estimates the engine state by an engine model.

In recent years, environmental regulations for ship propulsion plants have been tightened more and more, and CO ₂ reduction regulations represented by EEDI (Energy Efficiency Design Index) have also been tightened. The propulsion plant requires various countermeasure devices to comply with these regulations, and the system is complicated. However, as a propulsion plant, it is required to achieve both safety and further pursuit of energy-saving operation. For this reason, advanced actual sea area automatic application control technology for propulsion plants is required.
Some of the real sea area automatic application control technologies control the propulsion plant by creating a virtual plant with a propulsion plant simulator and comparing it with the actual plant. The core of the propulsion plant is the engine, and the engine model must be a model that faithfully represents the actual engine and is easy to calculate so that real-time control can be performed.

The most important state parameters in engine performance are the amount of fuel and the amount of air required for its combustion, and the excess air ratio (air-fuel ratio) is used to improve engine efficiency and reduce pollutants contained in exhaust gas. ) Is required to be properly controlled. However, conventionally, none of the parameters can be accurately measured especially in the unsteady state and cannot be used for control.

Further, in order to maintain the engine performance at the maximum efficiency, it is necessary to keep the engine state parameter at the optimum value in the given operating state. There are two conventional control methods for controlling engine performance, one is open-loop control and the other is closed-loop control.
Open-loop control is a method in which the relationship between a control action and a state parameter is created in advance with a steady-state map and used for control. Open-loop control is simple, but it cannot be applied to engine performance control that requires complicated control because the map cannot be applied when the engine state changes due to deterioration over time.
Closed-loop control is currently the most used. Closed loop control is a control that measures the engine state parameter (most simply, the engine speed) and minimizes the difference from the optimum set value. However, although PID control, which can be said to be a representative of conventional closed-loop control, is linear control, the actual engine state is non-linear, and the method using linear control is not optimal. In order to make up for this, it is necessary to create a map in advance and use it according to the engine state as in the case of open loop control, but in order to create a map, calculate the parameters according to all operating conditions. It takes a huge amount of time because it needs to be kept. Moreover, it cannot be guaranteed that the created map covers the non-linearity peculiar to the engine. Further complicating the problem is the reliability of the measurement sensor, which may include erroneous measurement due to deterioration and a lot of noise even if it can be measured, so that highly accurate control cannot be performed.

Here, in Patent Document 1, the values of parameters such as temperature acquired by the sensor of the engine are input to the state estimator on which the extended Kalman filter (EKF) is mounted, and the values of the input parameters and the prediction model of the engine are input. Use to estimate the state of the engine by estimating non-measured or non-sensing parameters, use an optimization algorithm to generate commands for actuators based on the state, and send the commands to the engine to improve engine performance. Methods for optimization are disclosed.
Further, Patent Document 2 includes a plurality of cylinders connected to the manifold, a detector for estimating the air-fuel ratio downstream of the manifold, and an estimator on which an extended Kalman filter (EKF) is mounted. A method of estimating the air-fuel ratio of each cylinder by inputting the delay time and the air-fuel ratio determined by the number and the angle of the crankshaft into the estimator is disclosed.
Further, Patent Document 3 discloses a method of measuring the crank angle of the crankshaft of an engine, calculating the crank angular velocity estimation error and the crank angle estimation error, and estimating the engine torque using a non-linear Kalman filter.
Further, in Patent Document 4, a combustion chamber is provided for a predetermined sealing period of the combustion chamber (after the exhaust valve is closed until the intake valve is opened, according to the oil / water temperature and the air supply pressure that change depending on the external environment and the like. A valve timing control device that corrects the closed period), the intake valve closing timing, the fuel injection amount during the sealing period, and the like is disclosed.

Japanese Unexamined Patent Publication No. 2005-248946 Japanese Unexamined Patent Publication No. 2006-336645 JP-A-2017-82662 JP-A-2002-129991

However, a technology that grasps the engine state by accurately estimating the excess air rate, which is the most important state parameter in engine performance, even in the unsteady state, and efficiently controls the engine based on the estimated excess air rate. Has not been proposed so far.

Therefore, according to the present invention, the engine performance can be maintained at maximum efficiency even in a non-steady state by estimating the excess air ratio using an engine model and controlling the engine based on the estimated excess air ratio. An object of the present invention is to provide an engine control method, an engine control program, and an engine control device using an observer.

In the engine control method using the engine state observer corresponding to the first aspect, engine control is performed by using an engine state observer that estimates the engine state by an engine model for an engine provided with an exhaust valve and fuel adjusting means. It is a method that sets the initial state value for calculating the amount of air to the engine, detects the rotation speed of the engine that drives the propeller of the ship as a sensor value, and estimates the load considering the disturbance in the actual sea area. Input to the engine model of the engine state observer linked with the propeller model, calculate the fuel amount based on the engine speed and load in the engine model, and based on the observation parameters of the engine model obtained from the initial state value and sensor value. Obtain the amount of air to the engine, estimate the excess air rate from the amount of fuel and the amount of air to the engine, and recover the excess air rate when the estimated excess air rate falls below the value at the beginning of operation. It is characterized in that at least the exhaust valve opening degree of the exhaust valve is controlled.
According to the first aspect of the present invention, since the excess air rate is estimated based on the engine speed or the like that can be measured reliably and accurately even in the unsteady state, the accuracy of the estimated excess air rate is high. Further, since the exhaust valve opening degree of the exhaust valve is controlled by estimating the excess air ratio, which is important for grasping the engine state, it becomes easy to maintain the engine performance at the maximum efficiency. Here, the excess air ratio is preferably an accurately estimated true excess air ratio. The true excess air ratio is the scavenging efficiency of the engine in the engine cylinder estimated in more detail based on the data measured by the target engine and the result of computational fluid dynamics (CFD) calculation. The excess air rate.

The present invention according to claim 2 sets the engine rotation speed, the supercharger rotation speed of the supercharger, the scavenging pressure, the exhaust temperature, and the exhaust pressure as the initial state values, and sets the engine of the engine state observer. The load as an observation parameter is estimated by the engine model using the fuel pump rack value detected as the sensor value after inputting to the model, and the gain to be multiplied by the control signal of the exhaust valve opening is calculated from the estimated load to exhaust. It is characterized by controlling the valve opening degree.
According to the second aspect of the present invention, since the exhaust valve opening degree of the exhaust valve is controlled by estimating the engine load in addition to the excess air ratio, the engine performance is further maximized even in the unsteady state. It will be easier to maintain.

The present invention according to claim 3 is characterized in that, when the excess air ratio becomes lower than the value at the initial stage of operation, the excess air ratio is recovered by controlling the valve closing timing to be earlier as the control of the exhaust valve opening degree. ..
According to the third aspect of the present invention, the excess air ratio can be effectively recovered, and the engine performance can be easily maintained at the maximum efficiency.

The present invention according to claim 4 is characterized in that, when the excess air ratio becomes lower than the value at the initial stage of operation, the fuel adjusting means for adjusting the amount of fuel so as to recover the excess air ratio is further controlled.
According to the fourth aspect of the present invention, since the fuel adjusting means is also controlled in addition to the exhaust valve opening degree of the exhaust valve, it becomes easier to maintain the engine performance at the maximum efficiency.

According to the fifth aspect of the present invention, the amount of fuel adjusted by the fuel adjusting means and the lift amount and / or operation timing of the exhaust valve are detected as sensor values and input to the engine model of the engine state observer to cause excess air. It is characterized by being used for estimating the rate.
According to the fifth aspect of the present invention, the accuracy of estimating the excess air rate can be improved.

The present invention according to claim 6 uses an engine model using the initial state value, the engine speed detected and input at time k, the fuel pump rack value, and the exhaust valve opening degree or operation timing of the exhaust valve. It is characterized by estimating the excess air rate and the load.
According to the sixth aspect of the present invention, it is possible to improve the estimation accuracy of the excess air ratio and the load.

The present invention according to claim 7 controls the exhaust valve opening degree of the exhaust valve based on the estimation result of the excess air ratio, and detects and inputs the engine rotation speed, the fuel pump rack value, and the exhaust at time k + 1. Engine speed, turbocharger speed, scavenging pressure, exhaust temperature, as observation parameters of the engine model for estimating the excess air rate and load in the engine model using the exhaust valve opening or operation timing of the valve. And the exhaust pressure is updated.
According to the seventh aspect of the present invention, it is possible to improve the estimation accuracy of the excess air ratio and the load by using the latest observation parameters.

The present invention according to claim 8 updates the engine model parameter for setting a non-linear model as an engine model suitable for a specific engine when the estimation error of the estimation result of the observation parameter deviates from a predetermined allowable range. It is characterized in that the observation parameters are estimated based on the updated engine model parameters.
According to the eighth aspect of the present invention, it is possible to reduce the error and improve the estimation accuracy of the observation parameter by the engine model parameter adapted to the specific engine.

The present invention according to claim 9 is characterized in that an extended Kalman filter or an unsented Kalman filter is used for estimating the excess air ratio and the load in the engine state observer.
According to the ninth aspect of the present invention, the accuracy of estimating the excess air ratio and the load can be improved by using the non-linear Kalman filter.

The present invention according to claim 10 is characterized in that an extended Kalman filter or an unsented Kalman filter is selected based on a setting state of an initial state value.
According to the tenth aspect of the present invention, the excess air ratio and the load can be accurately estimated according to the state of the engine model.

The present invention according to claim 11 is characterized in that an extended Kalman filter or an unsented Kalman filter is selected in advance and set in an engine state observer.
According to the eleventh aspect of the present invention, the excess air ratio and the load can be estimated quickly.

The engine control program using the engine state observer corresponding to claim 12 is an engine control program that controls an engine provided with an exhaust valve and fuel adjusting means by using an engine state observer that estimates the engine state by an engine model. In addition, a state initial value acquisition step for acquiring the input state initial value for calculating the amount of air to the engine and a sensor value acquisition step for acquiring the engine rotation speed at time k as a sensor value in the computer. Then, the initial state value and the sensor value are input to the engine model linked with the propeller model that estimates the load considering the disturbance in the actual sea area, and the engine model calculates the fuel amount based on the engine speed and load, and the initial state. An air excess rate state estimation step that obtains the amount of air to the engine based on the observation parameters of the engine model obtained from the values and sensor values, and estimates at least the excess air rate based on the amount of fuel and the amount of air to the engine, and the excess air When the excess air rate becomes lower than the value at the initial stage of operation based on the estimation result of the rate, at least a control step for controlling the exhaust valve opening degree of the exhaust valve is executed so as to recover the excess air rate. And.
According to the twelfth aspect of the present invention, since the air excess rate is estimated based on the initial state value and the engine speed that can be reliably and accurately measured even in the unsteady state, the accuracy of the estimated air excess rate is high. Is high. Further, since the exhaust valve opening degree of the exhaust valve is controlled by estimating the excess air ratio, which is important for grasping the engine state, it becomes easy to maintain the engine performance at the maximum efficiency even in the unsteady state.

According to the thirteenth aspect of the present invention, the exhaust valve opening degree of the exhaust valve is controlled in the control step based on the estimation result of the excess air rate in the excess air rate estimation step, and the rotation of the engine acquired as a sensor value at time k + 1. Update the observation parameters of the engine model at time k for estimating the excess air rate and load in the engine model using the number, fuel pump rack value, exhaust valve opening or operation timing of the exhaust valve. It is characterized by having more steps.
According to the thirteenth aspect of the present invention, the estimation accuracy of the excess air ratio and the load can be improved by using the latest observation parameters.

The present invention according to claim 14 is for setting a non-linear model which is an engine model suitable for a specific engine when the estimation error of the estimation result of the observation parameter in the air excess rate estimation step deviates from a predetermined allowable range. The engine model parameter update step for updating the engine model parameter is further provided, and the observation parameter is estimated based on the updated engine model parameter.
According to the 14th aspect of the present invention, it is possible to reduce the error and improve the estimation accuracy of the observation parameter by the engine model parameter tailored to the specific engine.

The invention according to claim 15 is characterized in that an extended Kalman filter or an unsented Kalman filter is used in the excess air ratio estimation step.
According to the fifteenth aspect of the present invention, the accuracy of estimating the excess air ratio and the load can be improved by using the non-linear Kalman filter.

The present invention according to claim 16 further includes a selection step of selecting an extended Kalman filter or an unsented Kalman filter based on a setting state of an initial state value.
According to the 16th aspect of the present invention, the excess air ratio and the load can be accurately estimated according to the state of the engine model.

The engine control device using the engine state observer corresponding to claim 17 is an engine control device using an engine control program using the engine state observer, and is a computer including the engine state observer and an engine model. It is characterized by including an input means for inputting the initial value of the state, a sensor value acquisition means for acquiring the engine rotation speed as a sensor value, and at least a control means for controlling the exhaust valve opening degree of the exhaust valve.
According to the 17th aspect of the present invention, since the excess air rate is estimated based on the engine speed that can be reliably and accurately measured even in the unsteady state, the accuracy of the estimated excess air rate is high. Further, since the exhaust valve opening degree of the exhaust valve is controlled by estimating the excess air ratio, which is important for grasping the engine state, it becomes easy to maintain the engine performance at the maximum efficiency even in the unsteady state.

According to the engine control method using the engine state observer of the present invention, the excess air rate is estimated based on the engine speed or the like that can be measured reliably and accurately even in a non-steady state. Is high. Further, since the exhaust valve opening degree of the exhaust valve is controlled by estimating the excess air ratio, which is important for grasping the engine state, it becomes easy to maintain the engine performance at the maximum efficiency.

In addition, the engine speed, the supercharger speed, the scavenging pressure, the exhaust temperature, and the exhaust pressure are set as the initial state values and input to the engine model of the engine state observer, and then the sensor. When the load as an observation parameter is estimated by the engine model using the fuel pump rack value detected as the value, and the gain to be multiplied by the control signal of the exhaust valve opening is calculated from the estimated load to control the exhaust valve opening. In addition to the excess air rate, the engine load is also estimated to control the exhaust valve opening degree of the exhaust valve, so that it becomes easier to maintain the engine performance at maximum efficiency even in a non-steady state.

In addition, when the excess air rate falls below the value at the initial stage of operation, the excess air rate is effectively recovered when the excess air rate is recovered by controlling the valve closing timing as a control of the exhaust valve opening. It can be done, and it becomes easier to maintain the engine performance at maximum efficiency.

In addition to the exhaust valve opening of the exhaust valve, when further controlling the fuel adjusting means for adjusting the amount of fuel so as to recover the excess air ratio when the excess air ratio becomes lower than the value at the initial stage of operation. Since the fuel control means is also controlled, it becomes easier to maintain the engine performance at maximum efficiency.

In addition, when the amount of fuel adjusted by the fuel adjusting means and the lift amount and / or operation timing of the exhaust valve are detected as sensor values and input to the engine model of the engine state observer to be used for estimating the excess air rate. The accuracy of estimating the excess air rate can be improved.

In addition, using the initial state value, the engine speed detected and input at time k, the fuel pump rack value, the exhaust valve opening or operation timing of the exhaust valve, the excess air rate and load in the engine model. In the case of estimating, the excess air rate and the estimation accuracy of the load can be improved.

In addition, the exhaust valve opening of the exhaust valve is controlled based on the estimation result of the excess air ratio, and the engine rotation speed, the fuel pump rack value, the exhaust valve opening of the exhaust valve, or the exhaust valve opening that is detected and input at time k + 1 When updating the engine speed, turbocharger speed, scavenging pressure, exhaust temperature, and exhaust pressure as observation parameters of the engine model for estimating the excess air rate and load in the engine model using the operation timing. By using the latest observation parameters, the excess air rate and load estimation accuracy can be improved.

In addition, when the estimation error of the estimation result of the observation parameter deviates from the predetermined allowable range, the engine model parameter for setting the non-linear model as the engine model suitable for a specific engine is updated and based on the updated engine model parameter. When estimating the observation parameters, it is possible to reduce the error and improve the estimation accuracy of the observation parameters by using the engine model parameters tailored to a specific engine.

In addition, when an extended Kalman filter or an uncented Kalman filter is used for estimating the air excess rate and load in the engine state observer, a non-linear Kalman filter is used to improve the estimation accuracy of the air excess rate and load. be able to.

Further, when the extended Kalman filter or the unsented Kalman filter is selected based on the setting state of the initial state value, the excess air ratio and the load can be accurately estimated according to the state of the engine model.

Further, when the extended Kalman filter or the unsented Kalman filter is selected in advance and set in the engine state observer, the excess air rate and the load can be estimated quickly.

Further, according to the engine control program using the engine state observer of the present invention, the excess air ratio is estimated based on the initial state value and the engine speed that can be measured reliably and accurately even in the unsteady state. The accuracy of the excess air rate is high. Further, since the exhaust valve opening degree of the exhaust valve is controlled by estimating the excess air ratio, which is important for grasping the engine state, it becomes easy to maintain the engine performance at the maximum efficiency even in the unsteady state.

Further, based on the estimation result of the excess air rate in the excess air rate step, the exhaust valve opening degree of the exhaust valve is controlled in the control step, and the engine rotation speed acquired as the sensor value at time k + 1 and the fuel pump rack value are used. If the observation parameter update step for updating the observation parameter of the engine model at time k for estimating the excess air ratio and the load in the engine model by using the exhaust valve opening degree or the operation timing of the exhaust valve is further provided. By using the latest observation parameters, the excess air rate and load estimation accuracy can be improved.

In addition, when the estimation error of the estimation result of the observation parameter in the air excess rate estimation step deviates from the predetermined allowable range, the engine that updates the engine model parameter for setting the non-linear model that is the engine model suitable for a specific engine. When the model parameter update step is further provided and the observation parameter is estimated based on the updated engine model parameter, the error is reduced and the estimation accuracy of the observation parameter is improved by the engine model parameter tailored to a specific engine. Can be done.

Further, when an extended Kalman filter or an unsented Kalman filter is used in the air excess rate estimation step, the air excess rate and the load estimation accuracy can be improved by using the nonlinear Kalman filter.

In addition, if a selection step for selecting an extended Kalman filter or an unsented Kalman filter based on the setting status of the initial state value is further provided, the excess air ratio and the load can be accurately estimated according to the state of the engine model. It can be carried out.

Further, according to the engine control device using the engine state observer of the present invention, the excess air rate is estimated based on the engine speed that can be reliably and accurately measured even in the unsteady state. High accuracy. Further, since the exhaust valve opening degree of the exhaust valve is controlled by estimating the excess air ratio, which is important for grasping the engine state, it becomes easy to maintain the engine performance at the maximum efficiency even in the unsteady state.

Conceptual diagram of an engine control device using an engine state observer according to the present embodiment Conceptual diagram of control using the same engine state observer Configuration diagram showing the configuration and relationship of the engine, control unit and load Functional block diagram of the control system of the exhaust valve The figure which shows the relationship between the exhaust valve opening degree and an excess air ratio. Flowchart showing how to set the engine model parameters and select the Kalman filter Flowchart showing how to estimate the engine state Diagram showing the estimation error of the same engine state

The engine control method, the engine control program, and the engine control device using the engine state observer according to the embodiment of the present invention will be described below.

FIG. 1 is a conceptual diagram of an engine control device using an engine state observer according to the present embodiment.
The engine control device 10 is a simulator of a propulsion plant incorporating each module of a propeller model and an engine model in consideration of a disturbance (Actual Sea Condition) representing an actual sea area such as a wave or wind condition, and is a so-called virtual plant (Virtual). Create a Plant), compare it with the actual plant (Real Plant), modify the engine model parameters offline or online, improve the accuracy of the simulator, and perform fine-grained control over the engine.
The engine control device 10 controls an engine including an exhaust valve and fuel adjusting means by using an engine control program using the engine state observer 15. The engine is mounted on the ship and drives the propeller that propels the ship as a load.
The engine control device 10 includes a computer 11 including an engine state observer 15, an input means 12 for setting an initial state value including an engine rotation speed, an engine rotation speed, a fuel adjustment amount by the fuel adjustment means, and an exhaust valve. The sensor value acquisition means 13 for acquiring the lift amount and the operation timing of the exhaust valve, and the control output means 14 for controlling the exhaust valve and the fuel adjusting means (fuel pump rack) are provided. The input means 12 is a control panel, a mouse, a keyboard, a touch panel, or the like. The exhaust valve can be acquired by the sensor value acquisition means 13 individually or in combination with the timing of valve closing and valve opening and the valve lift amount.

FIG. 2 is a conceptual diagram of control using an engine state observer according to the present embodiment.
The engine control device 10 obtains the rotational speed n _e of the engine 1 which drives the propeller 2 in sensor value acquisition unit 13 of the detector and the like, and inputs to the engine state observer 15. Engine state observer 15 estimates the load q _p excess air ratio λ of the engine as the engine state by the engine model using rotational speed n _e entered. Then, the engine control unit 10, the optimum gain is calculated, and sends a control signal of updating the exhaust valve opening EVC (the closing timing) fuel pump rack h _p. The engine control unit 10, based on the transmitted signal, to adjust the exhaust valve opening EVC and the fuel pump rack h _p by the control output unit 14. The adjustment of the exhaust valve opening EVC may be performed by combining the valve opening timing, the valve lift amount, and the like, in addition to the exhaust valve closing timing. Also, adjustment of the fuel control means, in addition to the fuel pump rack h _p, the timing of the fuel pump, an electronic governor may be a fuel adjustment mechanism or the like.
Here, in estimating the excess air ratio λ, the scavenging efficiency of the engine 1 obtained in advance by experiments, CFD simulations, or the like is used. In particular, in the scavenging process of a two-stroke engine, a part of the total supplied air passes through the cylinder, and the remaining supply air remains in the cylinder even after the scavenging is completed. At this time, since the residual exhaust gas composed of the combustion gas of the previous cycle exists in the cylinder, the total amount of gas in the cylinder after scavenging is the sum of the residual air supply and the residual exhaust gas. The ratio of the mass of the residual air supply after the completion of scavenging to the total amount of gas in the cylinder is called scavenging efficiency, and is a value indicating the quality of the scavenging action. The scavenging efficiency represents the air supply concentration in the cylinder and is a value directly linked to the engine output, but the reaching level differs depending on the scavenging method of the engine 1. Therefore, the true excess air ratio λ cannot be derived unless the correct scavenging efficiency of the engine 1 is grasped, and the engine performance cannot be truly maintained at the maximum efficiency based on the excess air ratio λ.
Thus, input to the engine state observer 15 detects at least rotational speed n _e of the engine 1, the real excess air factor λ of the engine based on the scavenging efficiency as the engine state in the engine state observer 15 load q _p is estimated, and the exhaust valve and the fuel adjusting means are controlled as control targets based on the estimated true excess air ratio λ. Reliably and on the basis of the measured accurately can the rotational speed n _e of the engine 1, and to estimate the excess air ratio λ based on the scavenging efficiency, while reducing the number of parameters to be used for estimation, accurately excess air ratio λ estimate it can. In addition, since the exhaust valve and the fuel control means are controlled by estimating the excess air ratio λ, which is important for grasping the engine state, it becomes easier to maintain the performance of the engine 1 mounted on the ship at the maximum efficiency and operate it. ..
At least one of the lift amount (opening) or operation timing of the exhaust valve and the fuel adjustment amount by the fuel adjustment means are acquired by the sensor value acquisition means 13 and input to the engine state observer 15, and the excess air rate. By using it for estimating the engine state including λ, the estimation accuracy of the engine state can be improved.
Further, the engine state observer 15 estimates the engine state in consideration of the state of the propeller 2 by the propeller model, so that the estimation accuracy of the engine state can be further improved.

FIG. 3 is a configuration diagram showing the configurations and relationships of the engine, the control unit, and the load.
The engine 1 assumes a 2-stroke diesel engine, and has an engine body 3, a supercharger 4, an air supply pipe, an intercooler 5, an air sweep receiver 20, an exhaust valve 6, an exhaust receiver 7, an exhaust pipe, and a fuel adjusting means 16 ( It includes a fuel pump 16a, an electronic governor 16b, a fuel adjusting mechanism 16c) and a controller 17 (exhaust valve timing changing mechanism 17a, exhaust valve drive pump 17b).

The engine body 3 includes a cylinder 3a, a piston 3b, a crank 3c, and the like. In the engine body 3, air is supplied and scavenged to the cylinder 3a according to the reciprocating motion of the piston 3b in the cylinder 3a. Further, fuel is supplied into the cylinder 3a by the fuel adjusting means 16, ignition, combustion, and explosion are performed in the air-fuel mixture, and the energy drives the piston 3b. The driving force applied to the piston 3b is transmitted to the load via the crank 3c. The main load in the present embodiment is the propeller 2 for propulsion of a ship, and the load of the propeller 2 fluctuates due to disturbance due to wave wind or a change in ship speed, and an unsteady state of the engine 1 occurs.

The supercharger 4 includes a compressor C and a turbine T. In the supercharger 4, the turbine T is rotated at high speed by using the energy (temperature and pressure) of the exhaust gas from the engine main body 3, and the compressed air is driven by driving the compressor C by the rotational force of the turbine main body 3. Is supplied into the cylinder 3a of. That is, the supercharger 4 sends the compressed air to the engine body 3 and sucks and explodes the air-fuel mixture that exceeds the original supply air amount of the engine 1 to give an output exceeding the apparent displacement amount.

The air compressed in the supercharger 4 is sent to the air supply pipe via a diffuser or the like. An intercooler 5 is provided in the air supply pipe. The intercooler 5 intermediately cools the compressed air. The compressed air that has passed through the intercooler 5 is sent to the scavenging receiver 20 and stored. The compressed air stored in the scavenging receiver 20 is sent into the cylinder 3a of the engine body 3 from the scavenging port that is opened when the piston 3b of the engine body 3 is near the bottom dead center. The fuel adjusting means 16 includes a fuel pump 16a, an electronic governor 16b, and a fuel adjusting mechanism 16c. Electronic governor 16b receives the speed signal N _e indicating the rotational speed of the crankshaft 3c, and controls the drive timing of the fuel pump 16a. The fuel pump 16a injects fuel into the cylinder 3a at a desired timing under the control of the electronic governor 16b. Fuel is supplied from the fuel adjusting means 16 into the cylinder 3a of the engine body 3 to form an air-fuel mixture, which is compressed by the piston 3b and the air-fuel mixture is burned. A driving force is applied to the piston 3b by combustion.

The exhaust gas generated by the combustion in the engine body 3 is scavenged by the compressed air stored in the scavenging receiver 20 as the exhaust valve 6 is opened and sent to the exhaust receiver 7. The exhaust gas stored in the exhaust receiver 7 is guided by the turbine T of the supercharger 4 to give a rotational force and then exhausted.

In the present embodiment, the valve opening timing for changing the exhaust valve 6 from the closed state to the open state and the valve closing timing for changing the exhaust valve 6 from the open state to the closed state are controlled by the control unit 18. Specifically, the control unit 18 functions as an exhaust valve timing control device, and the operation timing and opening degree (lift amount) of the exhaust valve 6 are controlled by controlling the exhaust valve timing changing mechanism 17a. The exhaust valve timing change mechanism 17a utilizes the oil pressure generated by the exhaust valve drive pump 17b to adjust the pressure at which the exhaust valve 6 is opened by an electromagnetic relief valve provided in the hydraulic actuator that pushes the exhaust valve 6 to exhaust. Control the valve 6. The control of the exhaust valve 6 will be described later. The exhaust valve 6 may be controlled only by the operation timing, only by the opening degree (lift amount), or may be combined, but the control of only the operation timing simplifies the configuration of the exhaust valve timing changing mechanism 17a. It is preferable because it can be done.

Further, in the present embodiment, the scavenging bypass pipe 19 is connected to the scavenging receiver 20. The scavenging bypass pipe 19 is provided with a flow rate sensor 19a and an take-out valve 19b. By opening the take-out valve 19b, a part of the compressed air sent to the scavenging receiver 20 is sent to the air lubrication supply pipe (not shown). On the other hand, by closing the take-out valve 19b, the compressed air sent to the scavenging receiver 20 is not sent to the supply pipe. The opening degree of the take-out valve 19b is controlled by the control unit 18 by using the signal of the flow rate sensor 19a.

FIG. 4 shows a functional block diagram of the exhaust valve control system. The ship in the present embodiment is equipped with a friction resistance reducing device that reduces the frictional resistance of the ship by extracting pressurized air (scavenging air) from the vicinity of the turbocharger 4 and ejecting air bubbles around the hull to lubricate the air. ing. Even when scavenging air is taken out during this air lubrication, it also corresponds to an unsteady state.
Engine state observer 15, the rotation speed sensor 13a is a sensor value acquisition unit 13, the rotational speed _{n e,} respectively from the fuel control amount sensor 13b and the exhaust valve sensor 13c engine 1, the fuel pump rack _{h p} and the exhaust valve opening EVC Get information about. The rotation speed sensor 13a is provided in the engine body 3, the fuel adjustment amount sensor 13b is provided in the fuel adjustment means 16, and the exhaust valve sensor 13c is provided in the exhaust valve 6 or the exhaust valve timing changing mechanism 17a. Then, to detect the rotational speed _{n e} of the engine 1 by the rotation speed sensor 13a, the fuel control amount sensor 13b detects the fuel pump rack _{h p,} detects the exhaust valve opening EVC by the exhaust valve sensor 13c.
Engine state observer 15, the rotational speed n _e of the engine 1 input _by the engine model using fuel pump rack h _p and the exhaust valve opening EVC load q _p of the excess air ratio λ as the engine state engine 1 estimated, the optimum gain is calculated, and sends a control signal of updating the exhaust valve opening EVC and the fuel pump rack h _p to the control output unit 14.
The control output means 14 sends a control setting signal of the exhaust valve opening degree EVC to the control unit (exhaust valve timing control device) 18. The control unit (exhaust valve timing control device) 18 that has received the control setting signal of the exhaust valve opening EVC controls the timing and opening (lift amount) for opening and closing the exhaust valve 6. Specifically, the operation timing and opening degree (lift amount) of the exhaust valve 6 are adjusted by the exhaust valve timing changing mechanism 17a. The exhaust valve timing change mechanism 17a controls an electromagnetic relief valve provided in the hydraulic actuator that pushes the exhaust valve 6 in response to a control signal from the control unit (exhaust valve timing control device) 18, thereby controlling the exhaust valve 6 The operating timing and opening degree (lift amount) of the exhaust valve 6 are adjusted by adjusting the opening pressure.
Further, the control output unit 14 transmits the control setting signal of the fuel pump rack h _p relative to the fuel control means 16. Fuel control means 16 which receives the control setting signal of the fuel pump rack h _p adjusts the amount of fuel in response to the control setting signal.

In the present embodiment, at least the exhaust valve 6 is controlled according to the state of taking out the taken-out air for air lubrication as an unsteady state. The amount of scavenging air taken out is modified in relation to the amount of air ejected as bubbles required for air lubrication. That is, the amount of scavenging air (scavenging air) taken out from the scavenging bypass pipe 19 via the scavenging valve 19b is controlled.

The take-out rate of the take-out air is controlled by the take-out valve 19b. Here, the amount of air taken out from the scavenging air is detected by the flow rate sensor 19a provided in the scavenging bypass pipe 19 shown in FIG. 4, and the amount of air to the engine 1 is calculated from the rotation speed _ne of the engine 1. , The extraction rate of the extracted air can be obtained. Instead of using the flow rate sensor 19a, the amount of taken-out air taken out may be obtained based on the opening degree (displacement) of the take-out valve 19b.

The take-out ratio means, for example, the ratio of the amount of pressurized air (scavenging air) supplied from the supercharger 4 to the amount of taken-out air. At this time, the amount is preferably a mass flow rate, and it is preferable that the amount is a ratio that ensures the amount of scavenging air supplied to the engine body 3 even if other ratios are taken. In actual control, the amount of scavenging air supplied to the engine body 3 may be directly measured by the flow rate, or may be substituted by scavenging air pressure or the like.

On the other hand, the sweep pressure in the engine 1 decreases according to the extraction rate of the taken-out air, and the performance of the engine 1 deteriorates. In addition, the amount of harmful substances in the exhaust gas may increase. Therefore, the control unit 18 first controls the exhaust valve 6 of the engine 1 according to the amount of the extracted air. Specifically, the control unit 18 controls the closing timing of the exhaust valve 6 according to the intake air. As a result, the control unit 18 functions as a timing control means.
Engine state observer 15 estimates the load q _p of the excess air ratio λ as engine state engine 1 by the engine model using rotational speed n _e, the fuel pump rack h _p and the exhaust valve opening EVC of the engine 1, calculates the best gain for your air extraction, it sends a control setting signal of the exhaust valve opening EVC and the fuel pump rack h _p to the control output unit 14.
If there is a difference between the previously controlled valve closing timing of the exhaust valve 6 and the valve closing timing of the exhaust valve opening EVC, which is the control setting signal of the engine state observer 15, the valve closing timing is corrected according to the control setting signal.
In the above example, an example in which scavenging air is taken out as pressurized air for air lubrication is shown, but as pressurized air, air supply is taken out from between the supercharger 4 and the intercooler 5 and used, or from the exhaust receiver 7. The exhaust may be taken out and used. It is also possible to take out and use a combination of scavenging, air supply, and exhaust.

FIG. 5 is a diagram showing the relationship between the exhaust valve opening degree and the excess air ratio. FIG. 5A shows the initial valve opening (EVC_0), the first exhaust valve opening (EVC_1), and the valve closing timing of the second exhaust valve opening (EVC_2).
The excess air ratio λ decreases due to non-steady conditions in the actual sea area where the waves and wind conditions are poor, deterioration of the performance of the turbocharger, and the above-mentioned scavenging extraction state, but the exhaust valve opening EVC is in the initial state. If (EVC_0) is left as it is, the original optimum value of the excess air ratio λ cannot be recovered. Therefore, as shown in FIG. 5B, the engine control device 10 sets the exhaust valve opening EVC to the first exhaust valve opening (EVC_1) according to the air excess rate λ estimated by the engine state observer 15. ) Or the second exhaust valve opening degree (EVC_2), and control is performed so as to recover the original optimum value of the excess air ratio λ. As a result, the engine performance can be maintained at maximum efficiency.

Next, a method of estimating the engine state in the engine control program will be described with reference to FIGS. 6 to 8.
In the present embodiment, an extended Kalman filter (EKF) or an unsented Kalman filter (UKF) is used for estimating the engine state in the engine state observer 15.

The following equation (1) is a non-linear relational expression between the engine state parameter and the excess air ratio λ.
Here, lambda excess air ratio, _{G a} is the air volume, _{G f} is the fuel amount, EVC exhaust valve _{opening, n tc} the supercharger speed, _{p s} is the scavenging pressure, _{T e} is the exhaust temperature, P _e exhaust pressure, n _e is the engine speed, h _p is the fuel pump rack.

Further, the following equation (2) represents a nonlinear state-space equation of the engine dynamic model. Rpm _{n e} of the engine, which supercharger speed _{n tc,} the scavenging pressure _{p s,} exhaust temperature _{T e,} the state parameters of the exhaust pressure P _e is, the fuel pump rack _{h p,} the control updates the load _{q p} of the engine The state space equation of how it changes is shown.
_Here, A i, the _{B i} nonlinear _{function, S Ai,} S _Bi engine dynamics is the set value of the engine model parameters for a particular engine.

Further, the following equation (3) represents a nonlinear state space equation of the engine dynamic model when the engine load q _p is unknown, but for the unscented Kalman filter (UKF).

Further, the following equation (4) is an LTI (Linear time invariant) model for an extended Kalman filter (EKF). By using this linear model, the calculation speed can be increased.

FIG. 6 is a flowchart showing a method of setting engine model parameters and selecting a Kalman filter.
First, by setting the engine model parameters S _An and S _Bn for a specific engine, a non-linear model suitable for the specific engine is created (non-linear model creation step S1).
After the nonlinear model creation step S1, it is determined whether or not the created model is uncertain (determination step S2).
If it is determined in the determination step S2 that the created model is uncertain, an unsented Kalman filter (UKF) using the equation (3) is selected (UKF selection step S3).
If it is determined in the determination step S2 that the created model is certain, the extended Kalman filter (EKF) using the equation (4) is selected (EKF selection step S4).
In this way, an algorithm for selecting either the unsented Kalman filter (UKF) or the extended Kalman filter (EKF) depending on whether the engine model parameters are uncertain or not is created in advance offline.
As shown in FIG. 1, the set engine model parameters can be updated offline as needed as compared with the actual plant.
These offline steps and updates can also be done online.

FIG. 7 is a flowchart showing a method of estimating the engine state. In the flowchart of FIG. 7, the engine state is estimated online in real time at predetermined time intervals (eg, 200 Hz).
First, the rotation speed _{n e} of the engine, the supercharger rotational speed _{n tc,} the scavenging pressure _{p s,} exhaust temperature _{T e,} the load _{q p,} acquires the exhaust pressure _{P e} as the state initial value (state initial value obtaining step of the engine S10). By acquiring and setting the initial state value, the estimation accuracy of the engine state can be improved.
After the state initial value obtaining step S10, from the sensor value acquisition unit 13, the rotational speed n _e of the engine at time k (k), the fuel pump rack h _p (k), and obtains the exhaust valve opening EVC (k) (Sensor value acquisition step S11). It should be noted that only the engine speed _ne (k) may be acquired.
After the sensor value acquisition step S11, the rotational speed of the engine at the time k obtained by the sensor value obtaining step S11 n _e (k), the fuel pump rack h _p (k), and an exhaust valve opening EVC (k), the state Based on the initial state value acquired in the initial value acquisition step S10, the engine state observer 15 estimates the observation parameter as the engine state and estimates the excess air ratio λ (engine state estimation step S12). Estimation observed parameters, the rotational speed _{n e} of the engine, a supercharger speed _{n tc,} the scavenging pressure _{p s,} exhaust temperature _{T e,} the load _{q p,} exhaust pressure _{P e} of the engine. By using an extended Kalman filter (EKF) or an uncented Kalman filter (UKF) for estimating the engine state in the engine state observer 15, the estimation accuracy of the engine state can be improved. Which one to use follows the selection procedure described with reference to FIG. The extended Kalman filter (EKF) or the unsented Kalman filter (UKF) may be selected in advance and set in the engine state observer 15. In this case, the engine state can be estimated more quickly.

After engine condition estimation step S12, a fuel pump rack h _p, and as an input control based on the condition that the exhaust valve opening EVC is estimated, controls the fuel control means 16 and an exhaust valve (control step S13). As a result, the engine performance can be maintained at maximum efficiency.
After control step S13, the time k for the next predetermined time at a time k + rotational speed of the engine in _{1 n e (k + 1)} , the fuel pump rack _{h p} (k + 1), and obtains the exhaust valve opening EVC (k + 1) ( The next time sensor value acquisition step S14) returns to the engine state estimation step S12.
Further, after estimating the observation parameters in the engine state estimation step S12, the observation parameters of the engine model are updated based on the estimation results (observation parameter update step S15).
In this way, the engine state observer 15 repeatedly estimates the engine state and updates the observation parameters of the engine model.

Further, FIG. 8 is a diagram showing an estimation error of the engine state. In FIG. 8, the vertical axis represents the error and the horizontal axis represents the time. The dotted line indicates the allowable range of the engine state estimation error, and the solid line indicates the engine state error estimation value.
The engine control device 10 tracks the estimation error of the engine state by the engine state observer 15, and updates the engine model parameter when it deviates from the predetermined allowable range (engine model parameter update step S16).
Then, based on the updated engine model parameters, the engine state is estimated in the engine state estimation step S12. As a result, the error of the engine model parameter can be reduced, the error estimation accuracy of the engine state can be improved, and the reliability can be improved.
In addition, when updating the engine model parameters, various methods such as updating when the engine model parameters deviate from the predetermined allowable range multiple times, integrating the error in time, and updating when the predetermined conditions are reached, etc. Can be adopted.

As described above, as a means for solving the problem of the conventional control method using closed and open loop control, in the present embodiment, it depends on the state of the engine model without using many measurement sensors for control. Therefore, an observer using an unsented Kalman filter (UKF), which is a non-linear Kalman filter called a soft sensor, or an extended Kalman filter (EKF) is used.
It is significant that this mathematical model is made up of two parts. That is, a) a non-linear model that fits a particular engine and b) a linear model that is perfectly tied to that non-linear model. This makes it possible to effectively track the non-linear process of the engine using the engine state observer 15 with an uncented Kalman filter (UKF) or the engine state observer 15 with an extended Kalman filter (EKF). It becomes. Furthermore, it is possible to estimate the state inside the engine, which cannot normally be measured. Particularly load important true excess air ratio λ of the engine in engine performance q _{p (propeller} load) can estimate, can be used to input control. As a result, not only the current operating condition of the engine state can be estimated, but also the engine state in the near future can be accurately predicted, and optimal and safe driving becomes possible.
The engine state parameter may be another parameter or combination as long as the excess air ratio λ can be estimated accurately, and the observation parameter may be any other parameter or combination depending on the control target, control item, etc. You can choose.

The engine control method, engine control program, and engine control device using the engine state observer of the present invention can be suitably used for controlling the engine of a ship propulsion plant. In particular, the excess air ratio in the cylinder of the 2-stroke engine can be kept optimal, and the engine performance can be maintained at maximum efficiency. It can also be used to control any engine with an exhaust valve and fuel conditioning means that at least the excess air rate is required for control. For example, it can be used to control a gas engine that requires finer control.

1 Engine 2 Propeller 6 Exhaust valve 10 Engine control device 11 Computer 12 Input means 13 Sensor value acquisition means 14 Control output means 15 Engine state observer 16 Fuel adjustment means S3, S4 Select step S10 State initial value acquisition step S11 Sensor value acquisition step S12 Engine state estimation step S13 Control step S15 Observation parameter update step λ Excessive air rate q _p Engine load _ne Engine speed

Claims (17)

It is an engine control method that controls an engine equipped with an exhaust valve and fuel adjusting means by using an engine state observer that estimates the engine state by an engine model, and sets an initial state value for calculating the amount of air to the engine. It is set and input to the engine model of the engine state observer linked with the propeller model that detects the rotation speed of the engine that drives the propeller of the ship as a sensor value and estimates the load in consideration of the disturbance in the actual sea area. In the engine model, the fuel amount is obtained based on the rotation speed and the load of the engine, and the air amount to the engine is obtained based on the observation parameters of the engine model obtained from the initial state value and the sensor value. , The excess air rate is estimated from the amount of fuel and the amount of air to the engine, and when the estimated excess air rate becomes lower than the value at the initial stage of operation, at least the exhaust so as to recover the excess air rate. An engine control method using an engine state observer, which controls the opening degree of the exhaust valve of a valve. The engine speed, the turbocharger speed, the scavenging pressure, the exhaust temperature, and the exhaust pressure are set as the initial state values and input to the engine model of the engine state observer. Further, the load as the observation parameter is estimated by the engine model using the fuel pump rack value detected as the sensor value, and the gain to be multiplied by the control signal of the exhaust valve opening degree is calculated from the estimated load. The engine control method using the engine state observer according to claim 1, wherein the exhaust valve opening degree is controlled. Claim 1 or claim, wherein when the excess air ratio becomes lower than the value at the initial stage of operation, the valve closing timing is controlled to advance the valve closing timing as the control of the exhaust valve opening degree to recover the excess air ratio. An engine control method using the engine state observer according to 2. Claims 1 to 3 include further controlling the fuel adjusting means for adjusting the amount of fuel so as to recover the excess air ratio when the excess air ratio becomes lower than the value at the initial stage of operation. The engine control method using the engine state observer according to any one of the above items. The fuel amount adjusted by the fuel adjusting means, the lift amount of the exhaust valve and / or the operation timing are detected as the sensor values and input to the engine model of the engine state observer, and the excess air ratio. The engine control method using the engine state observer according to any one of claims 1 to 4, wherein the engine is used for estimation. In the engine model, the initial value of the state, the rotation speed of the engine detected and input at time k, the fuel pump rack value, the exhaust valve opening degree of the exhaust valve, or the operation timing are used. 2. Engine control using the engine state observer according to any one of claims 3 to 5, which cites claim 2 or claim 2, wherein the excess air ratio and the load are estimated. Method. The exhaust valve opening degree of the exhaust valve is controlled based on the estimation result of the excess air ratio, and the engine speed, the fuel pump rack value, and the exhaust valve are detected and input at time k + 1. The number of revolutions of the engine and the number of revolutions of the supercharger as the observation parameters of the engine model for estimating the excess air ratio and the load in the engine model using the exhaust valve opening degree or the operation timing. The engine control method using the engine state observer according to claim 6, wherein the scavenging pressure, the exhaust temperature, and the exhaust pressure are updated. When the estimation error of the estimation result of the observation parameter deviates from the predetermined allowable range, the engine model parameter for setting the non-linear model as the engine model suitable for the specific engine is updated, and the updated engine model parameter is updated. The engine control method using the engine state observer according to claim 6 or 7, wherein the observation parameters are estimated based on the above. The engine according to any one of claims 6 to 8, wherein an extended Kalman filter or an unsented Kalman filter is used for estimating the excess air ratio and the load in the engine state observer. Engine control method using a state observer. The engine control method using the engine state observer according to claim 9, wherein the extended Kalman filter or the unsented Kalman filter is selected based on the setting state of the initial state value. The engine control method using the engine state observer according to claim 9, wherein the extended Kalman filter or the unsented Kalman filter is selected in advance and set in the engine state observer. An engine control program that controls an engine equipped with an exhaust valve and fuel control means using an engine state observer that estimates the engine state using an engine model.
A state initial value acquisition step for acquiring an input state initial value for calculating the amount of air to the engine, and a sensor value acquisition step for acquiring the engine rotation speed at time k as a sensor value. , The state initial value and the sensor value are input to the engine model linked with the propeller model that estimates the load in consideration of the disturbance in the actual sea area, and the fuel is fueled based on the engine speed and the load in the engine model. The amount is determined, the amount of air to the engine is determined based on the observation parameters of the engine model obtained from the initial value of the state and the sensor value, and at least the air is determined by the amount of fuel and the amount of air to the engine. An air excess rate estimation step for estimating the excess rate, and at least the exhaust so as to recover the excess air rate when the excess air rate becomes lower than the value at the initial stage of operation based on the estimation result of the excess air rate. An engine control program using an engine state observer, which comprises executing a control step for controlling the exhaust valve opening degree of the valve. Based on the estimation result of the excess air rate in the excess air rate estimation step, the exhaust valve opening degree of the exhaust valve is controlled in the control step, and the rotation speed of the engine acquired as the sensor value at time k + 1. And the observation of the engine model at the time k for estimating the excess air ratio and the load in the engine model using the fuel pump rack value and the exhaust valve opening degree or the operation timing of the exhaust valve. The engine control program using the engine state observer according to claim 12, further comprising an observation parameter update step for updating parameters. When the estimation error of the estimation result of the observation parameter in the excess air rate estimation step deviates from the predetermined allowable range, the engine model parameter for setting the non-linear model which is the engine model suitable for the specific engine is updated. The engine control using the engine state observer according to claim 12 or 13, further comprising an engine model parameter updating step to estimate the observation parameter based on the updated engine model parameter. program. The engine control program using the engine state observer according to any one of claims 12 to 14, wherein an extended Kalman filter or an unsented Kalman filter is used in the excess air rate estimation step. The engine control program using the engine state observer according to claim 15, further comprising a selection step of selecting the extended Kalman filter or the unsented Kalman filter based on the setting state of the initial state value. .. An engine control device using an engine control program using the engine state observer according to any one of claims 12 to 16, wherein a computer including the engine state observer and the engine model. It is provided with an input means for inputting an initial state value, a sensor value acquisition means for acquiring the rotation speed of the engine as the sensor value, and a control means for controlling at least the exhaust valve opening degree of the exhaust valve. An engine control device using a characteristic engine state observer.