JPH10105418A - Timing managing device for engine control software - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce the scale of a control software program to be used in an engine control system and to correct the timing of a task start processing. SOLUTION: The timing managing device of an engine control softrare providing a real time operating system is provided with a basic timer 11 to be used in the engine control software and an event handler 12 controlling the transition state of the plural tasks by seetting and clearing an event flag as against the synchronization-system events of various kinds of cycles through the use of the output of the basic timer in addition to the asynchronization-system events from various kinds of sensors. The event handler 12 generates an event flag where a bit value is set in a low-order bit and the one obtained by inverting the bit value of the count value is set in the high-order bit in the counting of the basic timer, obtains the AND of the bit value which is set in the event flag with the previously set bit value and decides the timing for processing the task.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a vehicle engine control system using a control software program, and more particularly, to a reduction in the size of a timing management device for a control software program and a correction of the timing of a task activation process.

[0002]

2. Description of the Related Art A control software program of the above-mentioned engine control system is composed of a plurality of tasks, and a real-time operating system (RTOS) is used to effectively use these tasks. This real-time operating system responds to information coming from sensors such as measuring instruments in a short time,
A process for changing the state to a READY state, a WAIT state, or the like is performed. Although various timing controls exist in an engine control system of a vehicle, timing management is performed using the real-time operating system. The event flag is used in the timing management.
explain.

As timing control of a timer system, for example, a task for processing a cycle of 2 ms is tsk-2m, a task for processing a cycle of 4 ms is tsk-4m, and a cycle of 8 m
A task for performing the processing of s is provided as tsk-8m, and the like.
The above tasks tsk-2m, tsk-4m, tsk-8
m, etc., is set to an event flag of flg-2.
m, flg-4m, flg-8m, etc.

FIG. 16 shows event flags flg-2m, f
It is a figure explaining formation of lg-4m and flg-8m.
As shown in the figure, a 2 ms timer, a 4 ms timer, an 8 ms timer, etc. are provided for the basic timer, and each output is an event flag flg-2m, flg-4m, fl
g-8m, etc., and tasks tsk-2m, tsk-4
m, tsk-8m, etc., respectively.

The event flags flg-2m, flg
In the formation of -4m and flg-8m, it is necessary to prepare each timer to prepare each event flag and manage the timing of each task. When the number of tasks is large, the number of timers is also increased. Accordingly, there is a problem that the number of timers should be reduced to reduce the scale of the system. FIG.
7 is a task tsk-2m, tsk-4m, tsk-8
It is a figure which shows the start timing of m, etc. In the real-time operating system, as shown in the figure, the task of the synchronous interrupt of the timer system is periodically started. However, asynchronous interrupt tasks other than timer-based synchronous interrupts must also be processed. And the problem that the processing is delayed occurs.

[0006]

SUMMARY OF THE INVENTION Accordingly, in view of the above problems and problems, the present invention reduces the number of timers to reduce the scale, and prevents the processing delay of a timer system synchronous interrupt due to an asynchronous interrupt task. It is an object of the present invention to provide a timing management device for an engine control software program that can be used.

[0007]

According to the present invention, there is provided a real-time operating system for controlling a transition state of a plurality of tasks in a short time in response to information input from various sensors of an engine. An engine control software timing management device comprising: one basic timer used for the engine control software;
In addition to the asynchronous events from the various sensors, an event flag is set and cleared for various types of synchronous events using the output of the basic timer to control the transition states of the plurality of tasks. An event handler, wherein the event handler forms an event flag in which a bit value is set to a lower bit of the count of the basic timer, and an inverted bit value of the count value is set to an upper bit. The logical product of the bit value set in (1) and the bit value set in advance is determined, and the timing at which the task is to be processed is determined. By this means, it is not necessary to provide a timer for each of a plurality of tasks in the periodic system, and the size of the apparatus can be reduced.

[0008] The event handler corrects the timing so that, if the time to start the processing of the task whose timing has been determined is delayed, the next task activation processing is delayed by the delay time. By this means, it is possible to prevent the processing from becoming dense at a specific timing and causing a processing delay. The delay time is rounded to a predetermined unit. By this means, the processing can be simplified.

The delay time is averaged by taking n times. By this means, it is possible to correct a steady delay. If the delay time exceeds the cycle of task processing, the next task activation processing is invalidated. By this means, the processing can be simplified. If the delay time exceeds the cycle of task processing, an error flag is set. By this means, overload and overlapping of priorities can be checked.

When the delay time exceeds the cycle of task processing, the priority of the start processing of the own task is raised. By this means, a delay can be prevented. When the delay time exceeds the cycle of the task processing and the timing of the next task activation processing is corrected, the next task activation processing is accelerated. By this means, it is possible to reduce the obstacle to the next task processing.

[0011]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing an engine system of a vehicle according to the present invention. As shown in the figure, the engine system has an engine control computer that inputs output signals from various measuring instruments, and controls the engine using the computer.

FIG. 2 is a diagram schematically illustrating the engine control computer of FIG. As shown in this figure,
As various measuring instruments, for example, an input interface circuit for inputting output signals from a water temperature sensor, a battery, a vacuum sensor, an intake air temperature sensor, an O ₂ sensor, a knock sensor, and the like, and an A connected to an output of the input interface circuit A / D converter, another interface circuit for inputting output signals of a throttle position sensor, a starter, a distributor, an air conditioner, a speed sensor, a neutral start switch, etc., and an electric load signal, and an A / D converter and another A CPU having a memory connected to the interface circuit of
A constant voltage power supply for the CPU, and an output interface circuit connected to the CPU, and as an actuator connected to the output interface circuit, ISCV, # 1,
There are 2, 3, 4 injectors, circuit opening rues, igniters (including spark plugs) and warning lights.

FIG. 3 is a schematic diagram illustrating the engine control computer of FIG. As shown in the figure, in the CPU of the engine control computer, the operation processing of the input / output interface (I / O) circuit 1, the execution processing of the control software program 2 including a plurality of tasks, and the execution processing of the control software program -Time operating system 3 (RTOS) that controls the system
Are performed. In task management by a real-time operating system, timing management for managing all tasks by timing and priority is performed by an event flag.

FIG. 4 is a diagram for explaining the task state transition of FIG. As shown in the figure, the real-time operating system dispatches the READY
The task with the highest priority among the tasks in the state is moved to the RUN state. Further, the real-time operating system issues a preemption request to start another task having a higher priority than the currently executing task (READ).
When the state is shifted to the Y state), the processing of the task being executed is interrupted and the state is shifted to the READY state.

Each task in the WAIT state has an execution condition (= event flag state), and the real-time operating system shifts the task in which the condition is satisfied by the change in the event flag state to the READY state as shown in FIG. Release the wait (WAIT). During execution, the task issues its own execution conditions (= event flag state) to the real-time operating system. The real-time operating system checks the state of the event flag. If the condition is not satisfied, the real-time operating system shifts the task to the WAIT state as shown in FIG.

The following table describes the event flag function.

[0017]

[Table 1]

Further, an event handler is further provided in the CPU as described below. FIG. 5 is a diagram illustrating the program configuration of the event handler. As shown in the figure, the system call set is issued to the real-time operating system at the timing of occurrence of the event x. issues flg (x). With this issuance, the real-time operating system sets the event flag x as described above, and the task of the control software program described later activates the task waiting for the event flag x in the READY state. After startup, the system call clr Flg (x) is issued to clear the event flag x. That is, the system call set This is to prevent the task waiting for the event flag from being started except for the issuance of flg (x).

FIG. 6 is a diagram for explaining a program configuration of a task. As shown in the figure, the task tsk-1 has a system call wai. Notify flg (FLAG-1) to the real-time operating system, and enter the wait state for the event flag x. The task tsk-1 is started (resume execution) by the dispatch of the real-time operating system, and the execution is shifted to the application processing module of the next step. In this task, a priority, an ID, a task state, and the like are set as task information in order to contribute to the processing of the real-time operating system. Also, information on task tsk-1 and FLA
The definition of G-1 is stored in the external information file as follows:
Is set.

FIG. 7 is a view for explaining the program configuration of the external information file. As shown in this figure, the external information file contains initial values of task status,
Task stack area, priority, task function, task ID
The timing parameter definition is, for example, a timing parameter label FLAG-tsk-1 of tsk-1.
1 is set as a parameter of the event flag flg (x). As shown in FIG. 17, for example, the task ID every 2 msec is tsk-2m, and the task ID every 4 msec
D is tsk-4m, task ID every 8msec is tsk
-8 m etc. Also, an event x every 2 msec is represented as an event 2m, and an event flag flg
(X) is represented as flg (2m). Similarly, 4mse
The event x for each c is represented as an event 4m, and the event flag flg (x) is represented as flg (4m). Similarly, an event x every 8 msec is represented as an event 8m, and an event flag flg (x) is represented as flg (4m). The same applies hereinafter.

The priority of a task is defined as, for example, tsk-2m>tsk-4m>tsk-8m>. FIG. 8 is a timing chart for explaining an operation example from the initial state of each task when the power is turned on. As shown in this figure, the event 2m, 4m, 8m
Etc. occur simultaneously, and the event flag flg (2m), the event flag flg (4m), and the event flag flg
(8m), etc. are set. Thereby, tsk-2
m, tsk-4m, tsk-8m, etc. are in the READY state. Then, the event flag flg (2m), the event flag flg (4m), and the event flag flg (8
m), etc. are cleared. Dispatch puts tsk-2m having the highest priority in the RUN state. tsk-2m
Notifies the real-time operating system of waiting for the next event flag flg (2m). The real-time operating system puts tsk-2m in a waiting state, and dispatch puts the tsk-4m having the highest priority in the RUN state. tsk-4m indicates to the real-time operating system the next event flag flg (4
m) Notify wait. The real-time operating system waits for tsk-4m, and dispatch sets the tsk-8m having the highest priority to the RUN state. Hereinafter, the same procedure is repeated.

FIG. 9 shows an asynchronous event during execution of tsk-2m.
Tie for explaining an example in which the event a is generated and tsk-a is started.
It is a mining chart. As shown in this figure,
If there is an interrupt due to the occurrence of vent a, the event
Dora is the system call set issues flg (a).
At the same time, the processing of taka-2m is interrupted. Real-time audio
The operating system uses the event flag flg (a)
Is set to bring tsk-a into the READY state. That
After event handler is system call clr flg
(A) is issued. Real-time operating system
The system clears the event flag flg (a) and
End the process of the event handler. Dispatching puts tsk-a in the RUN state if tsk-a> tsk-2m. tsk-
a is the system call wai issues flg (a).
The real-time operating system uses tsk-a
Put in the waiting state. Dispatch runs tsk-2m
State and restart the process. tsk-a is the system code
Le wai issue flg (2m). The following is the same as FIG.
Take the following steps.

FIG. 10 is a timing chart for explaining another example in which an asynchronous event a occurs during execution of tsk-2m and tsk-a is started. The difference from FIG. 9 is that when tsk−a <tsk−2m, tsk−2m is restarted, and then tsk−m
-A is in the RUN state.

FIG. 11 shows a system call set of an event handler in the timer system. FIG. 9 is a diagram illustrating issuance of flg (x). The basic timer 11 shown in FIG.
A 6-bit timer count value C-tmr is generated.
The event handler 12 connected to the outputs of the basic timer 11 and various sensors of the engine forms an event flag using those outputs as events. The event flag is 3 in the case of a 32-bit macro computer.
It is set to 2 bits long. With this event flag,
Synchronous tasks tsk-2m, tsk-4m, tsk-
8m, etc., the transition processing from the waiting state of the asynchronous tasks a, b, etc. to the REDY state is performed. Hereinafter, the event flag will be described.

FIG. 12 is a diagram for explaining the event flags of FIG. As shown in FIG.
The counter value C-tmr of the basic timer is assigned to the lower 16 bits of the bit length, and the counter value C is assigned to the upper 16 bits. inversion counter value C obtained by inverting tmr Assign tmr. FIG. 13 is a diagram illustrating the operation of the event flag.

In step S1, the count value C is incremented every time the basic timer counts up. Enter tmr. In step S2, the event flag flg Set tmr to inverted C−tmr × 2 ¹⁶ + C−tmr.

In step S3, the event flag f
lg tmr and the timer count value K that the task wants to process Event flag flg from tmr The logical product (AND) with the flag pattern formed corresponding to the setting method for tmr is calculated as in the following equation. Inverted C-tmr × 2 ¹⁶ + C-tmr∩Inverted K-tmr ×
2 ¹⁶ + K-tmr In step S4, it is determined whether all the bits indicated by the flag pattern have been set. This judgment is "N
If "O", the process proceeds to step S6. It should be noted that there is only one counter value that satisfies the logical product expression for an arbitrary counter value.

In step S5, the above judgment is made that "YE
If "S", the corresponding task is changed from the waiting state to the READY state. That is, the start condition is satisfied. In step S6, the event flag flg Clear all bits of tmr and end. Therefore, according to the present invention, the activation condition of a plurality of tasks can be satisfied by one basic counter, and a large number of counters are not required unlike the related art, so that the scale of the apparatus can be reduced.

FIG. 14 shows a timer count value K at which task processing is to be performed when the processing is delayed by taking an example of a timer-based task tsk-4m due to interruption of another high-priority task. FIG. 15 is a flowchart illustrating a correction example of tmr, and FIG. 15 is a time chart illustrating the correction of FIG. In FIG. 14, in step S10, the end time C ′ of the application process tmr.

In step S11, the task processing delay time td is obtained as follows. td = C ' tmr-K tmr This td is rounded off and rounded to the unit of msec. In step S12, the timer count value at which the current task process is to be performed is corrected as follows.

K tmr = K tmr + td In step S13, the timer count value at which the next task processing is to be performed is set as follows. K tmr = K tmr + 4 msec In step S14, the event flag wait is set as described above.

It should be noted that other tasks are similarly corrected. According to the present invention, by performing the correction for shifting the next processing start time by an amount corresponding to the processing start time delayed,
Task scheduling changes automatically,
Processing becomes dense at a specific timing, and it is possible to prevent the processing from being further delayed. That is, since the delay time td is variable depending on the processing time of another task, td
A shift in the relative timing of sk-4m can be prevented.

Next, the timer count value K for performing the task processing by averaging the delay time td by n times is described. tmr may be corrected. This is because the first-order delay is ignored, and the one that is constantly delayed is preferentially corrected. Next, when the delay time td is, for example, tsk−4 m, when the delay time temporarily exceeds 4 msec, no correction is performed. Since the current process exceeds the next process, the next process is invalidated without correction and the next process is performed, thereby simplifying the process.

Next, if the delay time td temporarily exceeds 4 ms, an error flag is set. This is because it is necessary to check for overload and overlapping of high priority items. Next, when the delay time td temporarily exceeds 4 ms, the priority of the invoking task is raised. This is to prevent the delay time td from temporarily exceeding 4 ms. next,
When the delay time td temporarily exceeds 4 ms and the correction of the next processing is performed, the expression of step S13 in FIG. 14 is changed as follows to speed up the start-up.

K tmr = K tmr + 1 msec This is to prevent trouble from occurring in the next process.
The above process is an example, and is similarly applied to different periods of other timer systems.

[Brief description of the drawings]

FIG. 1 is a diagram showing an engine system of a vehicle according to the present invention.

FIG. 2 is a diagram schematically illustrating the engine control computer of FIG. 1;

FIG. 3 is a schematic diagram illustrating the engine control computer of FIG. 2;

FIG. 4 is a diagram for explaining a task state transition in FIG. 3;

FIG. 5 is a diagram illustrating a program configuration of an event handler.

FIG. 6 is a diagram illustrating a program configuration of a task.

FIG. 7 is a diagram illustrating a program configuration of an external information file.

FIG. 8 is a timing chart illustrating an operation example from the initial state of each task when the power is turned on.

FIG. 9 is a timing chart illustrating an example in which an asynchronous event a occurs during execution of tsk-2m and tsk-a is started.

FIG. 10 is a timing chart illustrating another example in which an asynchronous event a occurs during execution of tsk-2m and tsk-a is started.

FIG. 11 shows a system call set of an event handler in a timer system. FIG. 9 is a diagram illustrating issuance of flg (x).

FIG. 12 is a diagram illustrating the event flags of FIG. 11;

FIG. 13 is a diagram illustrating the operation of an event flag.

FIG. 14 shows an interrupt from another high-priority task.
Timer count value K at which task processing should be performed when processing is delayed, taking timer-system task tsk-4m as an example t
9 is a flowchart illustrating an example of correcting mr.

FIG. 15 is a time chart for explaining the correction of FIG. 14;

FIG. 16 shows event flags flg-2m and flg-4.
It is a figure explaining formation of m and flg-8m.

FIG. 17 shows tasks tsk-2m, tsk-4m, and tsk.
It is a figure which shows the starting timing of -8m, etc.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... I / O interface 2 ... Control software program 3 ... Real-time operating system 11 ... Basic timer 12 ... Event handler

Claims (8)

[Claims] 1. A timing management device for engine control software having a real-time operating system for controlling a transition state of a plurality of tasks in a short time in response to information input from various sensors of an engine, wherein the timing control device is used for the engine control software. One basic timer and an asynchronous event from the various sensors, and an event flag is set and cleared for a synchronous event of various cycles using the output of the basic timer. An event handler for controlling a transition state of the task, wherein the event handler sets a bit value to the lower bit of the count of the basic timer, and sets an inverted bit value of the count value to an upper bit. A flag is formed, and the bit value set in the event flag is set in advance. A timing management device for engine control software, which takes a logical product of the obtained bit value and a logical value to determine a timing at which a task is to be processed. 2. The method according to claim 1, wherein the event handler corrects the timing so as to delay the next task activation processing by the delay time when the time to start the processing of the task whose timing is determined is delayed. The timing management device for engine control software according to claim 1, wherein: 3. The timing management device for engine control software according to claim 2, wherein the delay time is rounded to a predetermined unit. 4. The timing management device for engine control software according to claim 2, wherein the delay time is taken n times and averaged. 5. The timing management device for engine control software according to claim 2, wherein when the delay time exceeds a cycle of task processing, the next task activation processing is invalidated. 6. The timing management device for engine control software according to claim 2, wherein an error flag is set when the delay time exceeds the cycle of task processing. 7. The engine control software timing management device according to claim 2, wherein when the delay time exceeds a task processing cycle, the priority of the start processing of the own task is increased. 8. The method according to claim 2, wherein when the delay time exceeds the cycle of the task processing and the timing of the next task activation processing is corrected, the next task activation processing is accelerated. 4. A timing management device for engine control software according to 4.