JP2014066165A - Engine control device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an engine control device capable of enhancing the responsiveness of an actuator.SOLUTION: The engine control device includes a multiple core processor having different cores in which an application part and a platform part are mounted for one control function, the application part serving for computing a control value for the actuator subject to the control of the control function, the platform part using the control value for outputting a control signal to the actuator. During at least a time from when the application part computes the control value to when the platform part acquires the control value, the multiple core processor temporarily moves one of the application part and the platform part to the core in which the other is mounted to pretend as if the application part and the platform part are mounted in the same core.

Description

  The present invention relates to an engine control apparatus, and more particularly to an engine control apparatus having a multi-core processor for engine control.

Conventionally, an improvement in the performance of a processor has been realized by an improvement in the clock frequency, but in recent years it has reached its limit due to the problem of power consumption (heat generation). There is a relationship represented by P≈α × f ³ (α: coefficient) between the power consumption P and the clock frequency f. According to this relationship, if the clock frequency is doubled, the power consumption is 8 times. Therefore, a multi-core having a low clock frequency is more advantageous in terms of power consumption and heat generation than a single core having a high clock frequency. . In recent years, research to utilize multiprocessors and multicore processors has been advanced in order to cope with higher engine control performance.

  For example, Patent Document 1 discloses a power management system for a vehicle electronic control device having a multiprocessor. In this publication, in order to reduce power consumption, a control operation divided for each function is assigned to the processor, and the processor to which the control operation corresponding to the function is assigned is started based on the start or stop state of the function. Or it is disclosed to stop.

Patent Document 2 discloses a computing resource allocation device in a multi-core processor system. This publication discloses that cache utilization efficiency is improved by assigning a plurality of tasks belonging to the same process to the same processor.
The applicant has recognized the following documents including the above-mentioned documents as related to the present invention.

JP 2007-125950 A International Publication No. 2010/093003 Japanese Patent Laid-Open No. 8-069444

  In recent years, with the advancement of electronic control in automobiles, the scale and development man-hours of control software in ECU development are increasing. Control software is being standardized to deal with the increasing scale and complexity of control software. Generally, the control software is a software platform (hereinafter referred to as PF) composed of application software (hereinafter referred to as an application) that executes operations for realizing the control function, and various I / O drivers and OSs associated with the CPU function itself. (Referred to as software). For example, the fuel injection control function includes an application that calculates an injection amount, an injection timing, and the like, and PF software that performs processing for outputting an energization signal to the injector in response to an instruction from the application.

  In the case of a conventional single core CPU, the application of the same control function and the PF software are naturally mounted on the same core. However, in the case of a multi-core CPU, even if the control function is the same, the application and the PF software may be mounted on different cores for the purpose of limiting the ROM capacity and balancing the processing load between the cores.

  Of the software standardization activities that have become active in recent years, the AUTOSAR specification stipulates that PF software such as I / O processing is implemented in only one core. In this case, there is a possibility that an application for a certain control function and PF software are mounted on different cores.

  When an application for one control function and PF software are implemented in different cores, function calls, variable sharing, etc. between the application and PF software straddle the core. Therefore, the overhead (time delay) becomes large compared with the case where it is mounted on the same core. For example, the time from when the energization instruction is issued to the PF software that performs the process for outputting the energization signal to the injector from the application for calculating the injection amount, the injection timing, etc. until the actual I / O processing is performed The delay increases. If the time delay increases, the intended engine control may not be realized, and fuel consumption and drivability may deteriorate. Therefore, it is necessary to reduce the time delay and satisfy the response of the actuator required by the control function.

  The present invention has been made to solve the above-described problems, and an engine capable of improving the responsiveness of an actuator when an application for one control function and PF software can exist in different cores. An object is to provide a control device.

In order to achieve the above object, a first invention is an engine control device having a multi-core processor and having an application part and a platform part for one control function mounted on different cores,
The application unit calculates a control value of an actuator that is a control target of the control function,
The platform unit outputs a control signal to the actuator using the control value,
The multi-core processor is configured such that the application unit and the platform unit are mounted in the same core at least after the application unit calculates a control value until the platform unit acquires the control value. And moving means for temporarily moving one of the application unit and the platform unit to a core on which the other is mounted.

The second invention is the first invention, wherein
The multi-core processor includes a first case in which the platform unit is moved to a core on which the application unit is mounted, and a second case in which the application unit is moved to a core on which the platform unit is mounted. Responsiveness determination means for determining which is more responsive,
The moving means temporarily moves the platform unit to a core on which the application unit is mounted when it is determined that the responsiveness is higher in the first case, and in the second case And a means for temporarily moving the application unit to a core on which the platform unit is mounted when it is determined that the responsiveness increases.

  According to the first invention, when the application part (application) and the platform part (PF software) for one control function can exist in different cores, either the application part or the platform part is It can be moved dynamically to the installed core. Since the application unit and the platform unit exist in the same core from when the application unit calculates the control value to when the platform unit acquires the control value, the time delay can be reduced. For this reason, according to 1st invention, the responsiveness of the actuator which a control function makes a control object can be improved.

  According to the second aspect, either one of the application unit and the platform unit can be dynamically moved to the core on which the other is mounted so that the responsiveness is further improved. For this reason, according to the second invention, it is possible to satisfy the response of the actuator required by the control function.

It is a conceptual block diagram for demonstrating the system configuration | structure which concerns on Embodiment 1 of this invention. It is a conceptual lineblock diagram of ECU22 which is a multi-core processor. 4 is a flowchart of a control routine executed by an ECU 22 in the system according to Embodiment 1 of the present invention.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In addition, the same code | symbol is attached | subjected to the element which is common in each figure, and the overlapping description is abbreviate | omitted.

Embodiment 1 FIG.
[System Configuration of Embodiment 1]
FIG. 1 is a conceptual configuration diagram for explaining a system configuration according to Embodiment 1 of the present invention. The system shown in FIG. 1 includes an internal combustion engine (hereinafter simply referred to as an engine) 10. The engine 10 includes one or more cylinders. Each cylinder is connected to an intake passage for taking air into the cylinder and an exhaust passage for discharging exhaust gas from the cylinder. The engine 10 according to Embodiment 1 is a supercharged engine having a turbine in the exhaust passage.

  The system according to the first embodiment includes an engine control device 20 for controlling the engine 10. The engine control device 20 includes various sensors and actuators, and an ECU (Electronic Control Unit) 22. The ECU 22 is an arithmetic processing unit including a storage circuit including a ROM, a RAM, an input / output circuit, and an arithmetic processing circuit, and is a multi-core processor for engine control in which a plurality of cores are enclosed in one processor package. . Various sensors and actuators are connected to the arithmetic processing circuit via an input / output circuit.

  On the input side of the ECU 22, a crank angle sensor 30 for detecting a crank angle and a crank angular speed, an accelerator opening sensor 32 for detecting an accelerator opening corresponding to an operation amount of an accelerator pedal of the vehicle, and a shift lever Various sensors for detecting the operating state of the engine 10 such as the state, the battery capacity, the brake negative pressure, and the air conditioner performance are connected (not shown). In addition, navigation information such as road surface information and signal information is also input to the ECU 22.

  On the output side of the ECU 22, an electronically controlled throttle valve 34 for adjusting the amount of air flowing through the intake passage connected to the cylinder, an injector 36 for injecting fuel into the cylinder, and an air-fuel mixture in the cylinder are ignited. An ignition plug 38 for operating the valve, a valve stop solenoid 40 for operating a valve stop device for stopping the intake / exhaust valve, a high-pressure fuel pump 42 for increasing the pressure of fuel supplied to the injector 36, and a bypass passage bypassing the turbine An electronically controlled waste gate valve 44, a variable valve device 46 that makes the intake / exhaust valve open / close timing and lift variable, and an electronically controlled automatic transmission (hereinafter referred to as ECT (Electronic)). Various actuators for controlling the operating state of the engine 10 such as an ECT solenoid 48 that operates a controllable transmission). Others are connected.

  The ECU 22 controls the operating state of the engine 10 by driving various actuators according to a predetermined program based on the outputs of the various sensors.

  FIG. 2 is a conceptual configuration diagram of the ECU 22 that is a multi-core processor. Although FIG. 2 shows a controller and two cores, the number of cores is not limited to this. The controller outputs various instruction signals (start / stop, clock frequency setting, calculation assignment) to each core. Control software that realizes one control function includes an application that calculates the control amount of an actuator to be controlled, and PF software that includes various I / O drivers and an OS associated with the CPU function itself. For example, the fuel injection control function includes an application that calculates an injection amount, an injection timing, and the like, and PF software that performs processing for outputting an energization signal to the injector 36 in accordance with an instruction from the application.

  The above-mentioned AUTOSAR specification stipulates that the PF software is mounted on only one core. In this case, the application related to one control function and the PF software may be mounted on different cores. When the application and the PF software are mounted on different cores, function calls and variable sharing between the application and the PF software straddle the cores. Therefore, the overhead (time delay) becomes large compared with the case where it is mounted on the same core. For example, in the example shown in FIG. 2, since the injection application of the fuel injection control function and the injector drive PF are mounted on the same core, the time delay is not a problem. On the other hand, since the valve stop application of the valve stop control function and the valve stop solenoid drive PF are mounted on different cores, there is a possibility that time delay becomes a problem. If the time delay increases, the intended engine control may not be realized, and problems such as fuel consumption and drivability deterioration may occur. Therefore, it is necessary to reduce the time delay and satisfy the actuator responsiveness required by the control function according to the operating state.

  Therefore, in the system according to the first embodiment of the present invention, when the application for one control function and the PF software are mounted on different cores, the control object is based on the information of the following reference 1 to reference 4. The application that calculates the control amount of the actuator and the PF software that outputs the control signal to the actuator using the control value are dynamically moved to the same core.

  By implementing the application and the PF software on the same core during the processing of the control function (at least from the time when the application calculates the control value until the PF software acquires the control value), Time delay can be reduced. As a result, the response of the actuator to be controlled can be improved, and the response of the actuator required by the control function can be satisfied.

(Standard 1: accelerator opening, engine speed)
A description will be given of dynamically moving the application and the PF software to the same core based on the accelerator opening information.
When the accelerator opening increases at a speed equal to or higher than a predetermined value, that is, during acceleration, it is necessary to control the throttle valve 34 with high responsiveness. Therefore, at the time of acceleration, with respect to the throttle control function, either the application for calculating the control value of the throttle valve 34 or the PF software for acquiring the control value and outputting the control signal to the throttle valve 34 is mounted on the other. The responsiveness of the throttle valve 34 is enhanced by temporarily moving to the existing core.

  Further, at the time of acceleration, it is necessary to control asynchronous injection and the accompanying ignition timing with high responsiveness. Therefore, at the time of acceleration, regarding the fuel injection control function, either one of an application for calculating the injection amount and the injection timing of the injector 36 and PF software for obtaining the injection amount and the injection timing and outputting a control signal to the injector 36, The other is temporarily moved to the core on which the other is mounted, and the responsiveness of the injector 36 is improved. Further, at the time of acceleration, with respect to the ignition control function, either the application for calculating the ignition timing of the ignition plug 38 or PF software for acquiring the ignition timing and outputting the control signal to the ignition plug 38 is mounted on the other. By temporarily moving to the existing core, the responsiveness of the spark plug 38 is enhanced.

  Further, it is necessary to increase the fuel supplied to the injector 36 to a predetermined fuel pressure simultaneously with acceleration. Therefore, during acceleration, the other implements either the application for calculating the control value of the high-pressure fuel pump 42 or the PF software that acquires the control value and outputs the control signal to the high-pressure fuel pump 42 for the fuel pressure control function. The responsiveness of the high-pressure fuel pump 42 is enhanced by temporarily moving the core to the core.

  In addition, it is necessary to close the waste gate valve 44 at the same time as acceleration to ensure a supercharging pressure. Therefore, at the time of acceleration, regarding the waste gate valve control function, an application for calculating the valve closing control value of the waste gate valve 44 and the PF software that obtains the valve closing control value and outputs the valve closing control signal to the waste gate valve 44. Either one is temporarily moved to the core on which the other is mounted, and the responsiveness of the waste gate valve 44 is improved.

  Conversely, when the accelerator opening decreases at a speed equal to or higher than a predetermined value or when the accelerator opening decreases to a predetermined value or lower, valve stop control is executed to quickly stop the intake and exhaust valves in synchronization with the fuel cut. The purpose of this is to prevent the deterioration by suppressing the flow of fresh air into the catalyst. Therefore, it is necessary to control the valve stop solenoid 40 without delay. Therefore, at the time of fuel cut execution, regarding the valve stop control function, an application for calculating the valve stop control value of the valve stop solenoid 40, and the PF for obtaining the valve stop control value and outputting the valve stop control signal to the valve stop solenoid 40 One of the software is temporarily moved to the core on which the other is mounted, and the responsiveness of the valve stop solenoid 40 is improved.

  The same effect can be obtained even if the accelerator opening described above is replaced with the engine speed. The engine speed can be calculated based on the output of the crank angle sensor.

(Standard 2: Navigation information)
Based on navigation information including road surface information, signal information, and the like, a description will be given of dynamically moving an application and PF software to the same core.
There is a case where fuel cut execution can be predicted from the navigation information such as a long downhill in front of the vehicle. When executing the valve stop control for quickly stopping the intake / exhaust valve at the time of the fuel cut, regarding the above-described valve stop control function, either the application or the PF software is temporarily placed in the core on which the other is mounted. To increase the responsiveness of the valve stop solenoid 40.

(Standard 3: Information for judging shift switching)
A description will be given of dynamically moving an application and PF software to the same core based on information for determining shift switching.
As information for determining shift switching, a shift lever state, an engine speed, a vehicle speed, and the like are used. Since high response is required for ECT shift switching, the drive response request of the ECT solenoid 48 is high. Therefore, in the case of an automatic transmission (AT) vehicle, an application for calculating the control value of the ECT solenoid 48 based on map information of a fuel consumption curve (determined by the vehicle speed and engine speed) for switching the shift, and the control value One of the PF software that obtains the control signal and outputs the control signal to the ECT solenoid 48 is temporarily moved to the core on which the other is mounted, and the responsiveness of the ECT solenoid 48 is improved. Similarly for ECTs other than the AT, the application and the PF software can be dynamically moved to the same core based on information for determining shift switching.

(Standard 4: Restart condition from idle stop)
A description will be given of dynamically moving an application and PF software to the same core based on a restart condition from an idle stop.
As a restart condition from the idle stop, battery capacity, brake negative pressure, accelerator opening, air conditioner performance, and the like are used as parameters. When the engine is restarted from the idling stop state, each actuator is required to have high responsiveness. Each actuator includes a throttle valve 34, an injector 36, a spark plug 38, a valve stop solenoid 40, a high pressure fuel pump 42, a waste gate valve 44, a VVT control valve, a starter motor, and the like. It is determined whether or not the parameter of the restart condition described above has reached a set value. If the parameter has reached the set value, it is determined that there is a restart request. When there is a restart request, regarding each control function for controlling each actuator described above, either one of the application or PF software is temporarily moved to the core on which the other is mounted, and the response of each actuator is increased. Increase.

(flowchart)
FIG. 3 is a flowchart of a control routine executed by the ECU 22 in order to realize the above-described operation. The routine shown in FIG. 3 is repeatedly executed during engine operation by the operating core or controller of the ECU 22 prior to starting the processing of each control function. In the routine shown in FIG. 3, first, the ECU 22 determines whether or not the target actuator controlled by the control function is in a state where high responsiveness is required (step S100). For example, at the time of acceleration described above, the injector 36 that is a target actuator of the fuel injection control function is in a state where high responsiveness is required. If the determination condition in step S100 is not satisfied, the processing is continued from the beginning of this routine.

  When the determination condition of step S100 is satisfied, the ECU 22 determines whether the application for the control function and the PF software are mounted on different cores (step S110). In the example of the fuel injection control function described above, it is determined whether or not the injection application for calculating the injection amount and the injection timing of the injector 36 and the injector driving PF software for outputting the energization signal to the injector 36 are mounted on different cores. judge. If the determination condition in step S110 is not satisfied, the processing is continued from the beginning of this routine.

  When the determination condition of step S110 is satisfied, the ECU 22 determines whether or not the response (start of driving) of the target actuator is likely not to be in time for the deadline (step S120). When the application for one control function and the PF software are installed in different cores, the calculation time (time, crank position) of the control value by the application passes the latest calculation time that defines the latest calculation time. The determination condition is satisfied. Preferably, the ECU 22 stores in advance a map that defines the relationship between the driving state and the latest calculation time, and acquires the latest calculation time according to the driving state.

  In the example of the fuel injection control function described above, the drive start timing of the injector 36 may not be in time for the calculated injection timing depending on the injection amount and the calculation timing of the injection timing. When the determination condition of step S120 is not satisfied, that is, when the response of the target actuator is in time for the deadline, there is no problem in the response of the target actuator even if the application and the PF software are mounted on different cores. In this case, processing is continued from the beginning of this routine.

  When the determination condition is satisfied in step S120, that is, when the response of the target actuator is not in time for the deadline, the ECU 22 moves the PF software to the core on which the application is mounted (first case), In the case where the application is moved to the core in which the PF software is mounted (second case), it is determined which one has higher responsiveness (step S130). Which of the first case and the second case increases the responsiveness is determined by comparing the processing load at that moment for each core. Then, it is determined that the direction of movement to the core having a lower processing load increases the responsiveness. As for the processing load, for example, when a priority set in software is high and a task whose execution time is longer than a predetermined time is already operating at that moment, it can be determined that the processing load is high.

  Preferably, if the priority of the process A to be moved (control function task) is higher than the priority of the process B operating at that time, the process B is interrupted when the process A is moved. Since it can be started, the processing load is determined taking into consideration the “priority of the task to be moved” and the “priority of the task operating in each core at that moment”. That is, a core with many processes with low priority and few processes with high priority is determined as a core with a low processing load. More preferably, when the performance (clock frequency) of each core is different, the processing load is determined in consideration thereof.

  When the determination condition is satisfied in step S130, that is, when the responsiveness is higher in the case where the PF software is moved to the core where the application is mounted (first case), the ECU 22 mounts the PF software in the application. It is moved to the core that has been made (step S140). On the other hand, if the determination condition is not satisfied in step S130, that is, if the application is moved to the core on which the PF software is mounted (second case), the responsiveness is improved, the ECU 22 sets the application to PF Move to the core where the software is installed (step S150).

  After the process of step S140 or step S150, the process of the control function for driving the target actuator is started. Note that the moved application or PF software is returned to the original core as soon as execution is completed (as soon as high response is not required for the target actuator) by another routine.

  As described above, according to the routine shown in FIG. 3, when the application for one control function and the PF software can exist in different cores, the application and the PF software are based on various information of the engine. Either one can be dynamically moved to the core on which the other is mounted. Thus, since the application and the PF software exist in the same core from when the application calculates the control value until the PF software acquires the control value, the time delay can be reduced. Furthermore, according to the routine shown in FIG. 3, the application or PF software is moved to a core with a low processing load. Thereby, the responsiveness of the target actuator can be further increased, and the responsiveness required for the target actuator can be satisfied.

(Modification)
In the description of the routine shown in FIG. 3 described above, the injector 36 having the fuel injection control function is exemplified as the target actuator. However, the target actuator is not limited to this. The target actuator may be an ignition control function, a valve stop control function, a fuel pressure control function, a waste gate valve control function, or the like described above.

  Further, the present invention is not limited to an in-cylinder direct injection engine and a port injection engine, and can be applied. Further, the present invention is not limited to a spark ignition type engine and a compression self-ignition type engine, and is applicable. Further, the present invention is not limited to a supercharged engine and a naturally aspirated engine, and is applicable.

  The present invention can also be applied to the case where one core is assigned to various processes for one cylinder. In order to control an actuator such as an injector 36 provided for each cylinder, it is conceivable to install a fuel injection control function application and PF software for each cylinder in each core. Also in this case, the PF software may be mounted only on the representative core as common to each cylinder, and the PF software may be moved to each core in accordance with the operation timing of the injector based on the crank angle or the like. According to this, it is possible to ensure the response of the actuator while reducing the memory usage.

In the first embodiment, the ECU 22 is the “multi-core processor” in the first invention, the application is the “application unit” in the first invention, and the PF software is the first application. Each corresponds to a “platform portion” in the invention.
Here, the ECU 22 executes the process of step S140 or step S150 so that the “moving means” in the first or second invention executes the process of step S130. The “responsiveness determination means” in the invention is realized.

10 Engine 20 Engine control device 22 ECU
30 Crank angle sensor 32 Accelerator opening sensor 34 Throttle valve 36 Injector 38 Spark plug 40 Valve stop solenoid 42 High pressure fuel pump 44 Waste gate valve 46 Variable valve device 48 ECT solenoid

Claims (2)

In an engine control device having a multi-core processor and having an application part and a platform part for one control function mounted on different cores,
The application unit calculates a control value of an actuator that is a control target of the control function,
The platform unit outputs a control signal to the actuator using the control value,
The multi-core processor is configured such that the application unit and the platform unit are mounted in the same core at least after the application unit calculates a control value until the platform unit acquires the control value. Moving means for temporarily moving one of the application unit and the platform unit to a core on which the other is mounted;
An engine control device comprising: The multi-core processor includes a first case in which the platform unit is moved to a core on which the application unit is mounted, and a second case in which the application unit is moved to a core on which the platform unit is mounted. Responsiveness determination means for determining which is more responsive,
The moving means temporarily moves the platform unit to a core on which the application unit is mounted when it is determined that the responsiveness is higher in the first case, and in the second case Further comprising means for temporarily moving the application unit to a core on which the platform unit is mounted when it is determined that the response is improved.
The engine control apparatus according to claim 1.

Images