Note: Descriptions are shown in the official language in which they were submitted.
CA 02723610 2010-11-04
WO 2009/138558 PCT/F12009/050379
I
METHOD FOR DETERMINING DEAD ZONE OF VALVE
BACKGROUND OF THE INVENTION
[0001] The invention relates to a method for determining the dead
zone of a proportional control valve used for controlling an actuator of a
rock
drilling device, the method comprising controlling the control valve with an
electric P controller and measuring the position of the actuator in a
coordinate
system representing its movement with a separate position detector.
[0002] In a hydraulic rock drilling apparatus, actuators as feed beam
cylinders of booms are controlled by using proportional directional control
valves. In practice, this involves controlling the actuator in both
directions,
whereby, when the valve is in the middle position, i.e. in the 0 position, the
actuator does not move, and when the valve spool is displaced in either
direction, the actuator moves at a proportional speed relative to the spool
offset in such a way that the more the spool deviates from the 0 position, the
faster the actuator operates.
[0003] However, in present control systems it has been noted that
when the valve is operating, there is typically what is called a dead zone on
both sides of the 0 point where no pressure fluid flows to the actuator in
spite
of a control signal.
[0004] If this dead zone of the valve is known, control can be
corrected by taking the dead zone into account and by forming in an electric
control signal, for example, an initial offset value of the extent of the dead
zone
when starting from the 0 position, so that the initial value of the control
signal is
equal to the required error.
[0005] In practice this is not feasible, however, because the dead
zone varies valve-specifically. Further, the dead zone may be different in
different operating directions of the valve. Different valves also have dead
zones of different extents, so one single value to correct this matter does
not
exist.
BRIEF DESCRIPTION OF THE INVENTION
[0006] An object of this invention is to provide a method with which
the operation of a hydraulic actuator of a rock drilling device can be made
more accurately controlled than before.
[0007] The method according to the invention is characterized by
the method comprising the steps of:
CA 02723610 2010-11-04
WO 2009/138558 PCT/F12009/050379
2
(a) setting for the P controller, for controlling the control valve (8),
the first value of the controller gain to be 0 or such a value that the
actuator is
not expected to move due to the effect thereof, and setting for the actuator a
target value deviating from its current value, whereby the difference between
the current value and the target value of the actuator determines the first
difference value;
(b) adding to the value of the controller gain a value increment of the
extent of a predetermined step for the P controller for controlling the
control
valve;
(c) waiting for a predetermined time to establish whether the
position value given by the position detector of the actuator expresses that
the
actuator has begun to move;
(d) if, on the basis of the position value of the position detector, the
actuator has begun to move, moving to step (f);
(e) repeating steps (b) to (c) until, on the basis of step (d), it is time
to move to step (f);
(f) waiting until the movement of the actuator has, on the basis of
the position detector of the actuator, stopped, and determining a second
difference value as a difference value between the target position and the
position of the moment of stopping;
(g) determining at least one offset value in accordance with steps
(h) and/or (i) and storing it in memory to be used for compensating for the
dead
zone of the control valve when controlling the control valve;
(h) determining a first offset value as a product of the P controller
gain according to step (d) and the first difference value according to step
(a);
(i) determining a second offset value as a product of the P controller
gain according to step (d) and the second difference value according to step
(f).
[0008] An essential idea of the invention is to determine for each
control valve at least one control offset value corresponding to the dead zone
in the middle of the movement range of the control valve spool and, each time
the valve is controlled, taking into account the dead zone in the direction of
motion of the valve. One embodiment of the invention comprises determining
for each control valve in both directions of motion at least one control
offset
value corresponding to the dead zone and controlling the valve in both
CA 02723610 2010-11-04
WO 2009/138558 PCT/F12009/050379
3
directions of motion by using the corresponding offset value to compensate for
the dead zone of the valve.
[0009] An advantage of the invention is that a simple method
enables definition of the properties of each valve and of the valve-specific
dead zone in both directions of motion of the valve, if required, in the
control
system, whereby, when controlling the actuator, the dead zone of the valve
can be virtually eliminated irrespective of whether the error is purely due to
valve properties or caused by the valve properties and the control electronics
together.
BRIEF DESCRIPTION OF THE INVENTION
[0010] The invention will now be described in greater detail in the
attached drawings, in which
Figure 1 shows schematically an actuator of a rock drilling device;
Figure 2 shows schematically an exemplary characteristic curve of a
proportional directional control valve and the value of the control signal
required by it;
Figure 3 shows schematically a block diagram of a solution
applicable to the control of a valve and an actuator;
Figure 4 shows schematically a flow chart for applying the method
according to the invention.
[0011] For the sake of clarity, embodiments of the invention are
shown simplified in the figures. Similar parts are denoted with the same
reference numerals.
DETAILED DESCRIPTION OF THE INVENTION
[0012] Figure 1 shows schematically an actuator in a rock drilling
apparatus. Reference numeral 1 indicates schematically a base to which a
boom 3 is turnably connected by means of a joint 2. The boom 3 can be turned
around the axis of the joint 2 with an actuator, i.e. a hydraulic cylinder 4,
which
is connected by its one end to the base I with joints 5, and correspondingly
by
its other end turnably to the boom 3 with a joint 6. In order to control the
hydraulic cylinder 4, pressure medium from a pump 7 can be fed into it via a
proportional control valve 8, whereby from the other end of the hydraulic
cylinder 4, pressure fluid flows into a pressure fluid container 9 also via
the
valve 8. In order to control the hydraulic cylinder 4, the proportional
control
valve 8 is controlled with a control unit 10, to which also sensors 11
indicating
CA 02723610 2010-11-04
WO 2009/138558 PCT/F12009/050379
4
the turning angle of the boom are connected. Further, the figure shows a feed
beam 12 at the end of the boom 3, along which a drilling machine 13 and a
drill
rod 14 connected thereto move when a hole is being drilled in rock 15. Various
turning cylinders or rotary motors for turning the feed beam relative to the
boom, feed apparatuses for moving the drilling machine and other actuators
are generally known as such and since they do not essentially relate to the
invention, there is no need to show them in the drawings or explain them in
the
description in more detail.
[0013] Figure 2 shows schematically a typical characteristic curve of
a valve, where straight line A indicates a theoretical characteristic curve of
the
valve, i.e. the relation of the control signal arriving at the valve and the
amount
of flow of the pressure fluid passing through the valve. Correspondingly,
straight lines B1 and B2 show real control signals of the valve operation and
the volume flows of the pressure fluid passing through the valve.
[0014] As seen from the figure, the relation of the control signal and
the volume flow passing through the valve is such that when the control signal
is 0, the volume flow is also 0, and depending on the value of the control
signal, the volume flow through the valve changes linearly from zero in one
direction or the other.
[0015] In practice, there is a dead zone in the operation of the valve
either as a property of the valve or due to the combined effect of the valve
property and control electronics, this dead zone being +-20 units in Figure 2.
Therefore, volume flow through the valve does not begin until the value of the
control signal is 20 units in one direction or the other, and only above this
value
does the volume flow of the pressure fluid follow the value of the control
signal
through the valve.
[0016] If the situation were this ideal with all control valves, this
would be easily corrected by setting, in the control unit, the offset required
by
the dead zone in the control signal of the valve. In practice, this is not
feasible
because the dead zone of each valve and of the direction of motion of each
valve spool is individual.
[0017] Figure 3 shows schematically a block diagram of a solution
applicable to controlling a valve actuator. The figure shows a controller 16,
which is a part of a control unit or which may be formed by a computer
program or a part of it in the control unit. The controller 16 is connected to
control an entity 17 formed by a control valve 8 and an actuator with a
control
CA 02723610 2010-11-04
WO 2009/138558 PCT/F12009/050379
signal which is proportional to the difference value between the measured
position of the actuator and the desired position and which is here generally
referred to as control. The controller 16 is what is called a P controller,
i.e. it
controls the valve and via that the actuator in such a way that it is directly
proportional to the control signal arriving at it. Further, the figure shows
schematically a summing unit 18, to which an offset value and a control signal
given by the controller 16 are brought to control the valve. On the basis of
these, final control is generated for the control valve in the summing unit
18.
[0018] Calibration of the valve and actuator operation is performed
by first setting for the controller the gain value 0 or such a value with
which the
control valve does not yet begin to let pressure fluid to the actuator. At
this
stage, the offset value is naturally 0. After this, a target value is set for
the
actuator to be controlled, i.e. a position value deviating from the current
position, so that the difference value between the current position and the
desired position of the actuator will be a given value, for example in
calibration
a turning angle of 1 or for linear movement 1 m. Naturally, also other
numerical values relative to the angle or linear movement are feasible. The
gain value of the P controller gain may also be a value other than 0, as far
as
the control provided by the product of the gain value and the difference value
does not generate actuator movement. In practice, the gain value and the
target value for determining the difference value can be set in any order or
even simultaneously. Thus, in step (a) of claim 1, the intention is not to
have
determination of the values in the particular order mentioned but the setting
order may be any of the above-mentioned orders.
[0019] When the P controller gain is 0, the valve control is 0. This is
due to the valve control being, at each moment of time, the controller gain
multiplied by the difference value of the actuator. When the controller gain
is
sufficiently low, the actuator does not begin to move because the product of
the gain and the difference value does not exceed the control gain required by
the valve.
[0020] Next, the value of the P controller gain is increased. For
example value 1 may be set as gain, whereby P controller gain x difference
value, for example 1, is obtained as the value of the valve control.
[0021] At this stage, one must wait for some time, for example 0.5
seconds or, depending on the device, longer in order to see on the basis of
the
signal arriving from the position detection sensor whether this control fed to
the
CA 02723610 2010-11-04
WO 2009/138558 PCT/F12009/050379
6
control valve generates actuator movement, which would mean that the
position of the actuator begins to move towards the target position. If such
movement is not detected, P controller gain is increased with predetermined
increments, i.e. steps, one step at a time, and after each increment, there is
a
waiting period in order to detect possible movement of the actuator before the
P controller gain can be further increased. When movement is detected in the
actuator, the first offset value is the same as the controller gain multiplied
by
the above-mentioned difference value of the initial position and the target
position of the actuator.
[0022] If offset in the actuator is detected as early as after the
setting of the first gain value, there is no dead zone in the control valve,
and
the volume flow of the pressure fluid generating from the valve control may,
even with this minimum control, be too great to achieve a good control result.
[0023] On the other hand, when the actuator has started moving,
the first offset value according to the product of the gain value, which has
generated the movement and has been set for the P controller, and the
difference value is stored in memory, and it may be later used for controlling
the valve and thus the actuator.
[0024] When the actuator has started to operate as a result of the
control valve being opened, the valve control starts to decrease because its
value is the product of the P controller gain and the remaining difference
value.
The difference value is here the distance between the target position and each
particular momentary position. When the actuator continues the movement, the
value of the valve control is reduced, and at some point, the actuator stops
moving because the value of the control valve control is reduced with the
reduction of the difference value to such an extent that the control valve
stops
letting pressure fluid to the actuator and closes. The movement of the
actuator
may, of course, stop because the actuator achieves the target position, the
difference value being thus 0. The second offset value of the control valve,
at
which the closing took place, is obtained by multiplying the gain set for the
P
controller by that difference value of the actuator which was the offset at
the
end of the movement, and this second offset value is stored in memory. This
product of the P controller and the difference value, obtained when the
actuator movement stops, is the second offset value relative to the control
valve control and represents the dead zone in the middle area of the valve in
the valve operation. In reality, this gives both the first offset value of the
valve
CA 02723610 2010-11-04
WO 2009/138558 PCT/F12009/050379
7
control, representing the offset required for starting the valve control, and
the
second offset value of the valve control, representing the offset that
remained
when the valve control ended. These two values differ from each other, and
when the valve is controlled later, both values, only one value or some value
between them can be used for compensating for the dead zone. Likewise,
since these values are usually different in different directions of motion of
the
valve spool, it is possible to use only one offset value in both directions of
motion or a separate offset value in each direction. In embodiments requiring
extreme accuracy, both offset values obtained may naturally be used in both
directions of motion.
[0025] This determination of the dead zone of the control valve is
shown schematically in Figure 4, which illustrates an example of determining
the offset value, i.e. the dead zone, as a flow chart when both the first and
the
second offset value are determined.
[0026] In Figure 4 gain Kp, i.e. control value, of the P controller has
been set to 0 in the first step 21 of the flow chart, due to which the control
valve does not obtain control and therefore the actuator cannot move either.
[0027] Next, step 22 comprises setting position value 1 as reference
value Ref relative to the current position of the actuator, as well as the
desired
direction of motion. This value I may be for example 11' at a turning angle, 1
m
in linear movement or another suitable value expressed, by way of example,
as number 1.
[0028] Subsequently, in step 23, the current gain value Kp added by
one, i.e. Kp = Kp + 1, is set as the gain of the P controller .
Correspondingly,
one means here one predetermined extent of a step, which is expressed, by
way of example, with number 1.
[0029] This done, one must wait in step 24 as a consequence of the
gain value in question to see whether the value causes movement in the
actuator. This movement, if there is any, is detected on the basis of the
position value given by the position detector.
[0030] In step 25, a decision about how to proceed is made on the
basis of the position value given by the position detector and expressing
whether the actuator has started moving. If no movement is detected, one
returns to step 23, adding yet one predetermined gain step to the gain Kp of
the P controller. The loop of steps 23 to 25 is repeated as long as it is
observed in step 25 that the actuator has started moving. Then, in accordance
CA 02723610 2010-11-04
WO 2009/138558 PCT/F12009/050379
8
with step 26, the current control value formed by the product of the P
controller
gain value and the difference value Ref set in step 22, i.e. the first offset
value,
offset 1, is stored in memory to be used later for controlling the control
valve of
the actuator.
[0031] When actuator movement has been detected, one waits in
step 27 without making any changes to the control of the control valve.
[0032] In the next step 28, it is checked with the position detector of
the actuator whether the actuator is still moving, i.e. whether the position
value
given by the position detector changes. As long as this value continues to
change, one returns to step 27 and, in a predetermined time period, moves to
step 28 again to check whether the actuator is moving.
[0033] Once it has been observed in step 28 that the position value
of the position detector of the actuator does not change any longer, one moves
to step 29, where the difference value between the target position set for the
actuator and the position at the stopping moment, i.e. the difference value,
is
determined. Thus, the second offset value, offset 2, of the control valve, in
other words the extent of the dead zone in the middle area of the valve from
the 0 position towards the measured direction of motion, is obtained by
multiplying the value of the difference value by the gain value of the P
controller. This second offset value is stored in memory, and it may be used
in
schematic presentation according to Figure 3 by adding the obtained offset
value to the control value of the valve irrespective of whether the control
system used is analogue or digital, and also when using manual control, for
instance joystick control.
[0034] The above method may be implemented automatically in a
computer-controlled apparatus, in which case the control may be performed by
anyone and there is no need for control specialists. In rock drilling
apparatuses
having several different actuators, this calibration may be performed
automatically for all control valves and actuators controlled by them, in
which
case separate offset values are obtained for each of them. Control may also be
performed in both directions of motion, in which case the different offset
values
in different directions of motion can be stored for each valve and each
actuator. In some cases, sufficient control accuracy is obtained by selecting
the offset value from between the initial and final offset values and by using
this value for correcting the control. If desired, the same value can be used
in
both directions of motion.
CA 02723610 2010-11-04
WO 2009/138558 PCT/F12009/050379
9
[0035] The invention is explained only by way of example in the
description and drawings, and it is not in any way restricted to what is
presented therein. What is essential is that in the operation of a
proportional
control valve, the dead zone where the valve does not in reality control the
actuator is determined in the way described in the claims and that the error
value thus obtained is used for compensating for the dead zone of the valve
during the use of the device and the control of the valve. Once stored in
memory, the offset value or values may be used for controlling the actuator,
irrespective of the control manner used in real controlling or the mode of the
controller. Thus, the values may be used with P, D and/or l controllers as
well
as in control systems using both analogue and digital control signals for
controlling valves. Likewise, when manual controllers are used for
controlling,
the control system may take the offset values into account in a desired
manner.
[0036] In some cases, features of this application may be used as
such, irrespective of other features. On the other hand, features presented in
this application may be combined to form various combinations.
[0037] The drawings and the related description are only intended
to illustrate the idea of the invention. Details of the invention may vary
within
the scope of the claims.
