DE4217138A1 - Digital position controller for road vehicle throttles - has PID control actions summed with integral action limited and proportional plus derivative wide ranging - Google Patents

Abstract

The digital controller e.g. used to control the position of the throttle (17) of a road vehicle fuelling system, has an input from a foot pedal sensor (12) digitised (18) and compared (26) with the feedback from the throttle sensor loop (17, 20). The control action for adjusting the throttle is provided by proportional (38), integral (40) and derivative (32) elements. The outputs are summed (50) and used to control a pulse-width modulator (50) coupled to the drive stage (62). The proportional and integral terms operate on the position error, whereas the derivative responds directly to the throttle position. While the integral action is limited, the other two actions cover a wide range. USE/ADVANTAGE - Positional control of vehicle windscreen etc wipers, of throttle flaps and injection pumps. Rapid response. Accurate control.

Description

State of the art

The invention relates to a digital controller for vehicles as in Preamble of claim 1 shown.

Such controllers, in particular PD or PID controllers, i. H. Controller with Proportional, differential and / or integral part, are in Connection with vehicles or with control devices for one type Drive unit of a vehicle for position control of a power control elements (throttle valve or control rod of an injection pump), for Speed control, for regulating the exhaust gas recirculation rate, for regulation of the torque, etc., used.

For the sake of speed, accuracy and stability of the control loop have been in this context in the past in particular analog controllers are used. Such a controller is e.g. B. in connection with an idle control in DE-OS 30 39 435 (U.S. Patent 4,441,471). With the increasing use of Microcomputers in connection with the control of vehicles ge However, the use of digital controllers is becoming increasingly important.  

Especially in connection with actuators which are not linear Show behavior and / or are subject to sliding or static friction, he the representation of digital controls turned out to be difficult since it is desirable in terms of speed and in terms of overcoming the influence of friction the controller parameters (P-, I and D constants) large in terms of amount.

However, this can limit the stability of the rule circle and the controller in the event of large changes in amount its active control range, in which it has approximately linear behavior shows lead out.

It is therefore an object of the invention to provide a digital PID controller create, which also in connection with non-linear or rei satisfactory control results achieved.

This is achieved in that when calculating the P, D, and / or I-shares the sum of a limit value formed from these shares limit, while the shares are are limited, d. H. essentially to the limit of the numerical limit empire of the microcomputer can grow.

In general, the I component is advantageously set to the limit of The sum is limited, i.e. the I component always remains in active Re gel area. Otherwise the controller would not behave linearly.

Advantages of the invention

The digital controller according to the invention shows its rule cautiously advantageous properties. His reaction to setpoint Changes are quick, faults are quickly corrected.  

Through the measures according to the invention, the controller parameters can be can be chosen large in terms of load, so that in connection with friction adhered well to actuators whose static and / or sliding friction can be wounded. A satisfactory regulation of such Actuators can thus be reached.

There are special advantages when using this controller in Connection with a position control for a power control element ner internal combustion engine, such as a throttle valve or a control rod an injection pump or position controls for electromotive Clutch actuator and two-motor wiper systems etc.

Limiting the individual I component to the total value solves this Problem that if the I component z. B. by blocking the Actuator would grow beyond its active area Actuating element, since the I component is very slow, after loosening the Blocking takes the wrong position.

Further advantages result from the following description of Embodiments as well as from the dependent claims.

drawing

The invention is explained below with reference to the embodiments presented in the drawing Darge. Here, FIG. 1 shows a block diagram of the controller structure according to the invention. FIG. 2 shows time diagrams to clarify the control behavior of a controller while limiting the individual components, while FIG. 3 finally shows time diagrams in connection with the method of limiting the sum according to the invention. In FIG. 4 is a flow chart is shown as an advantageous embodiment of the digital controller.

Description of embodiments

Fig. 1 shows a block diagram of the digital controller according to the invention using the example of a PID controller in use as a position controller for the position of a throttle valve of an internal combustion engine in an electronic engine power control system.

In this case, a computing element or control element 10 is provided, which, among other things, supplies the input lines 12 and 14, respectively, which connect the computing element 10 with an operating element 16 that can be actuated by the driver or with the power control element 17 , the throttle valve. Both input lines are routed to analog / digital converters 18 and 20 , the output lines 22 and 24 of which lead to a first summing point 26 . The actual controller is designated by 36 .

In a preferred embodiment, line 30 leads from line 24 to differential portion 32 (D) of controller 34 . From the first sum point 26 , a line 28 leads on the one hand to the proportional component 38 (P) of the controller 36 , on the other hand to the integral component 40 (I) of the controller 36 .

In another advantageous exemplary embodiment, instead of the line 30 , the line 42 shown in broken lines can also be provided, which leads from the line 28 to the differential component 32 of the controller 36 .

The output line 44 of the differential component 32 , the output line 46 of the proportional component 38 and the output line 48 of the integral component 40 of the controller 36 via a limiting element 49 are brought together in a second summation point 50 . From the output line 52 of this summation point 50 leads to a dashed element 54 , from which the line 56 is guided to a Digi tal / analog converter 58 . The output line 60 of the digital / analog converter forms the output line of the computing element 10 and leads to an output stage 62 , the output line 64 of which is led to the power control element 17 .

Via the line 12 and the analog / digital converter 18 , the computing element 10 is supplied with a position value by the operating element 16 , which position value is detected by a position transmitter (not shown). This converts the A / D converter 18 into a digital value and subsequently converts it into the position setpoint for the controller. In a comparable manner, a position value of the power control element 17 is fed to the computing element 10 via the line 14 , which position value is detected by a position transmitter, also not shown. The converter 20 converts the actual value of the control into a digital value. At the summation point 26 , the setpoint / actual value comparison takes place to form the control deviation, which is output via line 28 to the controller 36 . It is provided in a preferred embodiment that only the actual value, the position of the power control element, is supplied to the differential component 32 .

Depending on the input signals, the controller components 32 , 38 and 40 form their input signals output signals (individual signal values) by differentiation, proportional amplification or integration, which are brought together in the sum point 50 to form the so-called PID sum (controller sum signal values). This sum is ultimately output via line 56 as a controller output signal to the digital / analog converter 58 and converted into an analog voltage signal. The ana log signal controls line 60, the output stage 62 , which actuates the power control element, not shown, via line 64 . Due to the control function, the position of the power control element 17 is set in such a way that it approaches or reaches the target position specified by the control element 16 and thus by the driver.

An analogous arrangement results in connection with speed control lungs, which depend on the target speed and actual speed ten, or in conjunction with other position control loops in the vehicle be used. Furthermore, the arrangement shown is also before partial in connection with any regulation in the closed rule Circle of operating sizes of internal combustion engines can be used, which especially work with friction-sensitive control elements.

There are also advantages in connection with a PD controller structure door.

The controller structure shown can lead to the problems mentioned above. In general, control engineering requires that the controller always works in an active control range in which it shows essentially linear behavior. Since, in particular in connection with friction-sensitive control elements and in the case of fast controllers, the controller parameters are advantageously large in terms of amount and the P, D and I components leave the active control range, element 54 is provided which limits the PID sum signal to limit values limited, which essentially designate the active rule area.

In an advantageous embodiment, the I component is based on the Total value of the controller limited as a single component.

This shows significant advantages over the possible limits tion of the individual P, D, and optionally I shares, since the this type of limit, as shown below, shares can reverse their effect with the opposite sign, so that only the I component works. When there is a large change in the setpoint the I component may also have a negative effect on the dynamics of the Control system, since the P and D components compensate each other and the I component determines and slows down the necessary control intervention.  

The time diagrams shown in FIGS . 2 and 3 serve to explain the advantages of the procedure according to the invention. Fig. 2 shows time profiles of the individual signals and the PID sum when the individual parts of the controller are limited, that is, a limitation device would be provided in lines 44 , 46 and 48 , respectively. Fig. 3 shows the same timings in the example of the procedure to the invention OF INVENTION does not mean if the individual components, but is limited by means of the PID sum of element 54.

FIG. 2a (3a) is pulled through the desired value progression, dashed actual-value curve, Fig. 2b (3b) of the timing of the propor tion alan part (38, 46), Fig. 2c (3c) the time course of the integral component (40, 48), Fig. 2d (3d) the time course of the Differentialan part ( 32 , 44 ) and Fig. 2e (3e) the time course of the PID sum ( 50 , 52 ).

At time t ₀ there is an abrupt change in the target value ( FIG. 2a, solid line). This means that the proportional component jumps to its upper limit in a jump due to the also abrupt control deviation (cf. Fig. 2a, solid and dashed line). Since the slower integral part and the differential part, which only depends on the actual value, retain the value 0 at time t ₀ , the sudden change in the proportional share according to FIG. 2b forms a sudden change in the PID sum according to FIG. 2e. Due to the resulting output signal, the actual value is slowly tracked to the setpoint. As long as the control deviation exceeds a certain amount, the proportional component according to FIG. 2b remains at its limit value, while due to the now occurring change in the actual value, the differential component according to FIG. 2d forms an output signal, which at a certain point in time t _{1 is} limited is coming. In the area of this point in time it can happen that in the course of the PID sum according to FIG. 2e an area arises in which the control output signal is 0. This leads to an unsatisfactory control behavior with a slowing, stopped or even moving in the opposite direction actuator. If the actual value approaches the setpoint, the proportional and differential components drop below their limits, causing the PID total to assume values other than 0.

In contrast, according to Fig., The unsatisfactory behavior improves rule Wesent Lich in unlimited individual shares and limited PID Total 3. The setpoint jump occurring at time t ₀ leads to an increase in the proportional component according to FIG. 3b, which under certain circumstances can go as far as to limit the number range of the microcomputer. As a result, the PID sum according to FIG. 3e is guided to its limit value, that is to say generates a maximum output signal, which leads to the power control element being tracked at the setpoint value at maximum speed. The change in the actual value over time after a certain time after the setpoint jump leads to an increase in the differential component, to a decrease in the proportional component. As the time period increases and the actual value rises more quickly, the differential component outweighs the proportional component, so that the differential component ultimately exceeds the proportional component as shown in FIG. 2d. As a result, the PID sum is brought to its lower limit value ( FIG. 3e), which leads to a maximum control intervention in the actual value speed reducing effect. This effectively limits the overshoot despite the strong, rapid control intervention. The differential portion rises again in the following according to FIG. 2d, as a result of which the PID sum, that is to say the control output signal, rises again and slowly regulates the actual value to the desired value.

The time behavior of the controller shown in FIG. 3e approximately represents an ideal PID controller which does not have the disadvantages described above. Satisfactory control behavior with high selected controller parameters is achieved, the control intervention taking place quickly and precisely without overshooting. This is particularly advantageous in connection with frictional actuators.

In FIG. 4, the digital controller according to the invention is shown in a flow chart form.

In step 100 , the setpoint and actual value are recorded, in step 102 the control deviation e (n) is calculated from the difference between the setpoint and actual value. The subsequent step 104 is used to calculate the proportional component P (n) of the current program run by multiplying the controller deviation e (n) by the predetermined proportional constant P_konst, which may be selected as a function of the operating variables. Then, in step 106, by multiplying the current control deviation e (n) by the integration constant I_konst, which can also be selected to vary depending on the operating conditions, the integral change component i (n) of the current program run is formed, which limits according to step 107 and im subsequent step 108 is added to the integral component I (n-1) of the previous program run. The current integral component I (n) is thereby formed. In the subsequent step 110 , the sum PI (n) of the proportional and integral part is finally formed.

The above constants are preferably dependent on the amount of the control deviation e (n) is selected, whereby with a large amount of Re gel deviation the constants due to the strong required then Rule intervention is greater in amount than in amount small control deviation.  

In step 112 , the actual differential component D (n) is formed by forming the difference between the actual values acquired from the current (ACTUAL (n)) and the one acquired in the previous program run (ACTUAL (n-1)) and multiplication with the differential constant D_const. In subsequent step 114 , the PID sum PID (n) is calculated by adding the differential component D (n) thus formed to the PI component PI (n). This sum is subjected to a limitation in step 116 and possibly limited to predetermined limit values which denote the active control range of the controller. Finally, in step 118, the limited total value is output as a control output signal.

A comparable procedure is also advantageous applied to so-called PD controllers, where the Integralan was partially waived.

The formation of the differential component can also be based on actual evaluate from several program runs under appropriate weight tion take place, whereby the current values are of greater importance.

It is advantageous if the sampling time of the controller, i. H. the time points of the program call at the actuating time of the control element in Be drawing is set. In a preferred embodiment the ratio of positioning time: sampling time advantageous to approx. 1: 30 to Set at 1:40.

Claims (8)

1. digital controller for vehicles,

• with a proportional, differential and preferably integral component,
• - With means for forming individual signal values for each of the controller components
• with means for forming a controller sum signal value from the calculated individual signal values of the proportional, differential and preferably integral component,
• with means for forming an output signal from the controller sum signal value for controlling an actuating element for influencing the variable to be regulated in the sense of a regulation,
  characterized in that
• - Means are provided which limit the controller sum signal value to a predetermined range of values, while the individual signal values, preferably P and D components, can grow beyond the predetermined range of values.

2. Digital controller according to claim 1, characterized in that it in conjunction with a position control the performance of a burner engine influencing actuator, such as a throttle valve or an injection pump, an electromotive clutch actuator or from two-motor wiper systems. 3. Digital controller according to one of the preceding claims, characterized characterized in that it is in connection with a speed, torque ten-, filling or air volume or mass control use finds. 4. Digital controller according to one of the preceding claims, characterized characterized that he in connection with frictional Stell elements with linear and non-linear behavior, like a choke valve or injection pump control element, is used. 5. Digital controller according to one of the preceding claims, characterized characterized that the range of values is the range of the active rule is range in which the controller is a substantially linear Ver hold has. 6. Digital controller according to one of the preceding claims, characterized characterized in that the proportional and preferably existing whose integral part is one of the difference between the setpoint and actual value formed control deviation is supplied and in its dependence the individual signal values are formed during the differential Proportion of the actual value, depending on which the on individual signal value of the differential component is formed. 7. Digital controller according to one of the preceding claims, characterized characterized in that the amount of proportional and / or integra tion constant depending on the amount of the control deviation is such that the constants be in the case of a larger control deviation are larger in terms of daily income.   8. Digital controller according to one of the preceding claims, characterized characterized in that the formation of the output signal at intervals takes place at a ratio of 1 to the actuating time of the actuating element 20 to 1 to 50, preferably 1 to 30 or 1 to 40.