IE55385B1 - Working depth control system for vehicle with ground working implement - Google Patents

Description

2 2 55385 This Invention relates to a working depth control system for a vehicle having connecting means for attaching a ground-penetrating Implement thereto and an actuator means for raising and lowering the implement in response to a control signal derived from a combined 5 signal formed from a plurality of error signals representing differences between actual parameter signals derived from a machine system such as draft force, driven shed slippage and tractor engine speed, said machine system comprising the vehicle, the attaching means and the Implement, and corresponding set-point signals.

Such a system is known from DE-A- 2 508 620.

Most currently available hydromechanical Implement control systems operate to control the working depth of the Implement purely as a function of the implement position, purely as a function of the draft force sensed in the hitch linkages or as a mixed function of 15 the sensed Implement position and sensed draft force. Such control systems are limited In that they do not prevent excessive wheel slippage nor do they prevent engine stalling during certain adverse conditions. For example, with draft sensing only, the conventional draft force responsive control system may cause the engine to be 20 lugged down and possibly stalled, due to inaccurate draft sensing, increased rolling resistance or burrowing of the drive wheels. It has been proposed as disclosed in US 2,927,649; US 3,776,322, US 3,834,481 and US 3,913,680 to control the implement working depth purely as a function of the amount of slippage of the drive wheels, 25 thereby using wheel slippage input as a substitute for a sensed 3 draft force Input. Of course, such systems do not provide the advantage of draft force control, nor do they necessarily prevent engine stalling. A rotary Implement control system, described In US 4,077,475, includes multiple independent control modes which are individually selected by a selector switch. These control modes include a sensed draft force only control mode, a sensed position only control mode, a manual control mode and a mixed control mode wherein the implement working depth is controlled in response to a mixture of hitch position and engine speed. Another control system is described in US 3,860,074, wherein the control signal is the most positive of a draft error voltage and a position error voltage.

GB 1 503 592 combines a position error signal or an actuator pressure signal selectively with a thresholded' slip signal and a limited engine torque signal.

The object of the present invention is to provide an improved system which can adapt better to' the complex dynamics of the machine sstem comprising the vehicle, attaching means and implement.

Accordingly, the present invention provides a working depth control system for a vehicle having connecting means for attaching a ground-penetrating implement thereto and an actuator means for raising and lowering the implement to vary the ground penetration thereof in response to a control signal derived from a combined signal formed from a plurality of error signals representing differences between actual parameter signals such as draft force, driven wheel slippage and tractor engine speed, said machine system derived from a machine system comprising the vehicle, the attaching means and the implement, and corresponding set-point signals, the combined signal being formed as a combination of at least three parameter error signals, each of which is the difference between a sensed parameter value and a corresponding set-point value, the sensed parameter values being engine speed, wheel slip and draft force and the engine speed set-point value and draft force set-point value being provided by operator controls .

An advantage of the control system is that it can operate over a variety of field conditions without requiring frequency operator adjustments. 4 In an embodiment of the present invention, the largest of the combined error signal and position error signal is selected and applied to the input of an electrohydraullc control valve which controls fluid flow to and from 5 a cylinder which raises or lowers the implement, thereby reducing the nagnitude of the selected error signal· Means are provided for aodifylng the engine speed and wheel slip error signals in response to predetermined operator-induced changes in a load command setting which represents a desired level of draft force. Means are also 10 provided so that various of the reference values are redetermined only upon expiration of a certain time period after the occurrence of the predetermined change in the load command setting.

Such a system prevents excessive wheel slippage of a tractor pulling an implement while also controlling the implement working 15 depth as a function of other.variables. The system counteracts implement lowering resulting from reduced travel speed.

A particular advantage is that the system provides for implement depth control simultaneously as a function of sensed draft force, sensed wheel slip and sensed engine speed.

The invention will be described in more detail by way of example and with reference to the accompanying drawings, in which: Fig 1 is a simplified schematic of an agricultural tractor equipped with the present invention.

Figs 2a-g are flow charts of the algorithm performed by the 25 central processing unit shown in Fig 1.

Fig 3 is a simplified schematic representation of a central processing unit suitable for use in the present invention. λ tractor 10 includes an engine 12 controlled by a speed control lever 14, a pair of driven wheels 16 and a pair of non-driven wheels 18. A rear housing 20 supports a rear axle 22 and a rockshaft 24. An implement hitch 26, such as a conventional 3 point hitch, includes draft links 28 which are connected to lift arms 30 via lift links 32. The lift arms 30 are connected to the rockshaft 24 to insure simultaneous and equal movement and are raised and lowered via a pair of parallel connected hydraulic lift or rockshaft cylinders 34. A drawbar 36 extends rearwardly from the housing 20. The tractor 10 and the hitch 26 are merely exemplary and those skilled in the art will understand that the Invention can be applied to tractors and' hitches of other configurations. For example, this invention can be used on an articulated four-wheel drive tractor or on a front-wheel drive row-crop tractor.

An integral-type, ground-engaging implement (not shown), such as a. moldboard plow or a chisel plow, may be' attached in a conventional manner to the draft links 28. Alternatively, a towed implement (not shown) may be coupled to the drawbar 36. Draft sensors 38 may be interposed in the hitch 26, for example, in the draft links 28, to sense the draft force transmitted to the draft links 28 from the integral implement. If a plurality of draft sensors are interposed in the hitch 26, then a single draft representative signal may be obtained by combining together the separate signals from the plurality of sensors 38. In the case of a towed implement,.the draft force may be sensed with a draft sensor interposed in the drawbar 36, or with a T-bar coupled to the draft links. In either case, any suitable known draft sensor would suffice.

The communication of hydraulic fluid to and from the cylinders 34 or to and from a remote cylinder (not shown) on a towed or semi-integral implement is controlled by a conventional solenoid-operated electrohydraulic flow control' valve 42 which receives electrical, control signals generated by a control -unit or 'central processing unit 50, which may include a digital microprocessor, analog-to-digital converters, multiplexers and/or other equivalent conventional hardware components. The flow control valve 42 may be comprised of a commercially available MOOG-type servo-valve with a torque motor-operated 6 -pilot stage and an integral second stage. The output hydraulic flow rate is substantially proportional to the magnitude of the electrical current applied to the torque motor of the valve 42.

An operator-controlled position command lever 52 is coupled 5 to a lever position transducer 54 (such as a potentiometer) which generates a position command signal which represents a desired hitch or implement position. Electrical upper and lower position reference or limit signals ace provided by operator-adjustable potentiometers 51 and 53. Alternatively, adjustable 10 mechanical stops or limits (not shown) could be provided to mechanically limit the position of-control lever 52 and thus, limit the signal from potentiometer 54.. λ manually-operable __ -load command knob 56 is coupled to a potentiometer-type load transducer 58 which generates a load -command signal representing 15 a desired level of draft force.

A position transducer 60, such as a conventional rotary potentiometer, generates a sensed position signal which represents the actual sensed position of the rockshaft. λ position feedback signal could also be obtained from the lift 20 cylinder 34 or from a remote lift cylinder if that cylinder includes a position transducer, such as described in D. S.

Patent No. 3,726,191, for example.

A conventional rotation speed sensor 62, such as a Motorola magnetic pickup, generates a wheel speed signal which represents 25 the rotation speed of the rear axle 22 and the driven wheels 16. In the case of the magnetic pickup, a signal is generated with a frequency proportional to the wheel speed. A ground speed sensor 64, such as a commercially available doppler radar- type unit, generates a ground speed signal which represents the 30 actual ground speed of the tractor 10. The radar ground speed signal also has a frequency proportional to the ground speed of » the vehicle. As an alternative, a fifth wheel or a rotation , speed sensor for a non-driven wheel could also be used for this purpose. Suitable conversion factors may be applied by the 35 central processing unit SO to these speed signals to convert them to comparable units.

A speed control lever position sensor 66, such as a potentiometer coupled to the speed control lever 14, generates a signal which represents a desired engine speed. Alternatively, 40 it would be possible to obtain a desired engine speed signal 7 from the setting of an engine governor. Another conventional rotation speed sensor 68. such as an instrumentation-type tach generator or a magnetic pickup, generates an engine speed signal with a frequency proportional to the actual rotation speed of 5 the engine 12.

An operator-controlled activity control 70 is coupled to a transducer 72, such as a potentiometer, for generating a variable gain signal which is communicated to the central processing unit SO. Also, an operator-controlled, rate-10 selecting switch 74 may be opened and closed by the vehicle operator to select slow and rapid rates of implement movement.

As shown in Fig. 3, in the preferred case where the centra^ processing unit 50 includes a digital microprocessor, a suitable central processing unit 50 could be comprised of a conventional 15 analog-to-digital convertor with multiplexer 300, a microprocessor 310 and a programmable interval timer (PIT) 320. The A/D convertor 300 would convert the analog signals from transducers 38, 51, 53, 54, 58, 60, 66 and 72 to digital representations for communication to the microprocessor 310.

The PIT 320, such as an AMD 9513, would convert the square-wave signals from speed sensors 62, 64 and 68 to digital representations, also for communication to the microprocessor 310. The microprocessor 310, such as an Intel 8751 or T.I. 9940, would generate digital HVCO values according to the 25 algorithm described in detail hereinafter. Finally, the PIT 320 would convert the digital HVCO values to pulse-width modulated valve driving signals for driving the conventional torque motor controlled valve 42 to retract and extend the cylinders 34 and ' raise and lower the implement (not shown) attached thereto. The 30 control unit 50 could be implemented with other herdware configurations as well. For example, the microprocessor 310 could include internal memory, or memory could be external to the microprocessor 310, or a latch could be inserted between the microprocessor 310 and the indicating devices 80 and 82.

Appropriate scaling factors are applied to the signals from all the sensors so that their minimum and maximum values are represented by digital counts of 0 and 255, respectively. In this manner, the system has the capability of responding to changes in the input variables as small as approximately 0.4% 40 and the variables can be stored in 8-bit registers. 8 Conventional EMI filters could be in various portions of the lines connected to the control unit 50 to filter out electromagnetic interference.

The sequence of calculations performed by the central processing unit 50 is described next, with reference to the flow charts shown in Figs. 2a - g and the following tables: CENTRAL PROCESSING UNIT INPUT T»BLE INPPT NAME Ail) Bitch position (from sensor 60) 10 A{2) Lower position limit (from sensor S3) A(3) Raise position limit (from sensor 51) A(4) Activity control (from sensor 72) A(5) Hitch position command (from sensor 54) A(6) Draft load command (from sensor 58) 15 A(7) .Speed control lever position (from sensor 66) A(8) Engine speed (from sensor 68) A(9) Ground speed (from sensor 64) A(10). Draft sensor (from sensor 38) A(11) ' Rear wheel speed (from sensor 62) 20 The flow charts shown in Figs. 2a - g include abbreviations and acronyms which are more fully described later, but which are initially defined as follows: COMM Interim combined command value DGAIN Dynamic gain value 25 DIFF Engine speed error differential DLXT Downshift light DRAFT Draft force (sensed) EERR Engine speed error ELIH Engine speed error limit 30 ENGS Engine speed FAETSL Implement movement rate flag FDBK Position feedback FDRFT Filtered draft force FLAG Load command change flag 35 FPOS Filtered position feedback FSLP Filtered slip GAIN Activity gain value BVCO Valve command LCOH Load command 40 LCONT Load control mode flag 5 9 10 15 20 25 30 35 LERR Load or combined error LHYS Load hysteresis LLIM Lower position limit LLIT Load light PCOH Position command PERR Position error PBYS Position hysteresis PUT Position light RCOH Rate limited position command RERR Rate limited engine speed error RSCL Rate limited speed control lever position SCI. Speed control lever position SERR Slip error SFPOS Stored filtered position feedback SUP Driven wheel slip SSET Slip set point STHRS Stored engine speed threshold THLL Engine speed threshold lower limit THOL Engine speed threshold upper limit THRS Threshold (engine speed) TIMER Load command change delay timer - TVEL Tractor velocity (ground speed) OUM Opper position limit WVEL Rear wheel velocity The major portions of the flow chart, shown in Figs. 2a - g, can be summarized as follows: The slip error value, SERR, is generated via steps 102 - 128. The filtered engine error value, PERK, is generated via steps 130 - 160. The position feedback and filtered position feedback values, FDBK.and FPOS, are generated via steps 170 and 180. λ sensed draft force value, DRAFT, is generated by step 182, A value representing a desired amount of draft force, LCOM, is generated by step 186. Via steps 194, 196, 119, 156 and 160, the wheel slip and engine speed error values are momentarily modified whenever the operator makes a predetermined change in the load command control 56. This permits the draft control operating points to be set in steps 224 - 236 without undue interference from the wheel slip and engine' speed error inputs to assure that the draft force responsive aspect of this control system is normally the primary or dominant control aspect. 40 10 The central processing unit 50 enters into its routine at a step S8r whereupon the following initial conditions are established at step 99.

DRAFT- 0 5 ENGS « 0 FDRFT- 0 FLAG - 0 FPOS « 0 FSLP 0 10 LCONT- 0 LEYS * 0 PCOM »0 PHYS - 0 ‘ RCOM - 0 15 RERR » 0 RSCL « 0 SSET * 30 STHRS» 45 TIBER- 0 20 TVEL « 0 WVEL - 0 The main processing loop is entered at step 100 and is followed by step 101, wherein the values A(l) - A(ll) are obtained or read from operating parameter sensors and from the 25 operator-controlled adjustments. It will be evident that various off-setting or compensating factors can be applied to these values representing the sensed parameters to compensate for errors in calibration or adustment of the sensors. Then, in step 102, a filtered tractor velocity (ground speed) value TVEL 30 is calculated from the statement TVEL · (3 x TVEL + A(9))/4, where the TVEL value on the right side is the initial value set in step 99 or the value established in step 102 of a previous cycle of the routine; where A(9) is the sensed ground speed value from sensor 64. Thus, the TVEL value is a weighted 35 average of a previous TVEL value and the current sensed ground speed value.

Next, at step 104, a scaled and filtered rear wheel velocity value WVEL is derived from the current rear wheel speed input value A(ll) from sensor 62 by the statement WVEL « (3 x WVEL + 40 A(ll))/4. At step 106, the rear wheel velocity, WVEL, is 11 compared to a reference value of 30, which is a digital nunber representing a speed of .3 meters per second. If WVEL is less than .3 meters per second, then the program proceeds to steps 112 and 114. If WVEL is not less than .3 meters per second, 5 then the routine goes to step 108 where a SLIP value is calculated from the statement SLIP * 100 (WVEC - TVEL)AVEL. Thus, the SLIP value varies between zero and 100, as a percentage figure. Following step 108, the SLIP value is compared to 0 in step 110. If the SLIP value is less than 0, 10 then SLIP is set equal to 0 at step 112. If the SLIP is not less than 0, then the routine proceeds to step 114. In this manner, the value of SLIP will be greater than or egual to zero for the subsequent calculations in the routine.

At step 114, a filtered slip value, FSLP, is derived from 15 the statement FSLP (127 x FSLP + SLIP1/128. Thus, the updated filtered FSLP value is a weighted average of the previous FSLP value (initially 0 or as set in step 114 of the previous cycle), and the most recently determined SLIP value 'from step 108 or 112. The weighting factors described here prevent rapid 20 fluctuations in the FSLP value in response to short duration fluctuations in the SLIP value.

Next, at step 116, a maximum value, representing, for example, 304 wheel slippage, is established as an upper limit for the slip set value, or reference slip value, SSET, initially % 25 30 or set by step 204 of a previous cycle of the routine. In this manner, the SSET value is limited to a maximum valde representing 301, but it can be less.

Next, at step 118, the slip error value SERR is calculated -from the statement SERR SLIP - SSET. Thus, the slip error 30 value, SERR, is the difference between the actual wheel slippage value, SLIP, and the slip reference value, SSET.

Next, at a portion of the routine represented by step 119, if a FLAG value is equal to 1 (as set in step 196 of a previous routine cycle), then the slip error value, SERR, is rate limited 35 down to a zero value'. More particularly, the SERR value is incrementally decreased during each cycle of the routine by an amount which increases uniformly so that the SEPR value will be decreased gradually without any significant discontinuities, and as a substantially parabolic function of time. The increments are chosen so that normally the rate limited SERR 40 12 value would be reduced to zero in a time period of the order of .2-.5 seconds.

Hext, at steps 120 - 128, a dual slip error gain is established, depending upon the magnitude of the slip error 5 value, SERB, as determined in step 118 or 119. For example, if, in step 120, SERB is less than zero, then the routine proceeds directly to step 130 and the unmodified SEBP value from step 118 is used in the remainder of the routine. However, if SERB is not less .than zero, then SERB is compared to a value such as 20 10 in step 122. If SEBB is greater than 20, then a new SEBB value is obtained from steps 124 and 126 so that the new SEBP value is equal to 1.5 x ((old SERB value x 2) - 20). However, if SEBB is not greater than 20 in step 122, then the routine proceeds directly to step 126 where the gain factor 1.5 is applied to the 15 SERR value. In this manner, the control system is able to respond more rapidly to larger magnitudes of wheel slippage represented by the SERB value. The values of the gain factors in steps 124 and 126 may be experimentally determined to achieve the desired sensitivity of the system to wheel slippage. The 20 particular values may vary, depending upon the particular vehicle and implement involved. Following step 126, an upper limit 255 is established for the SERR value to prevent overflow of the 8-digit register containing that value.

At step 130, the speed control lever position value, SCE, is 25 derived from the value A(7) from the speed control lever position transducer 66, so that, for example, 1 count of SCE represents 10 rpm of engine speed. Next, in a portion of the routine represented by step 132, a rate limited speed control lever value, RSCE, is derived from the SCE value in step 130.

Specifically, the RSCE value (initially set at 0 in step 99) is changed during each cycle of the routine by an amount which is proportional to changes in the speed control lever position as represented by changes in the values A(7) and SCE. If the speed control lever position is changing in the direction 35 corresponding to increasing engine speed, then the incremental ~ increase in the RSCL value during each cycle of the routine is ' limited so that the RSCE value can increase no faster than a rate corresponding to a rate of increased engine speed of, for example, 700 rpm per second. The RSCE value is not rate-limited 40 when the speed control lever is moved in the direction 13 representing a lowering of the engine speed· As will be seen from later steps in the routine, this rate limit on the RSCL value will, in the event of a rapid advance of the speed control lever, limit the engine speed error value, EERP, by limiting the 5 updated rate limited s^eed control value RSCL, and eventually prevent undesirable raising of the hitch in response to a rapid advance of the engine speed control lever.

At step 134, the updated engine speed value, ENGS, is calculated from the statement ENGS (3 x ENGS + A(8))/4, where 10 A(8) is the value from the engine speed sensor 68, and where the ENGS value on the right is initially zero from step 99 or is set at step 134 of a previous cycle of the routine. The ENGS value is scaled so that 1 count represents 10 rpm. -- Next, at step 140, an engine speed threshold value, TBPS, it 15 assigned the value of the stored engine speed threshold value, STHRS, which is initially 45, in step 99, corresponding to a deadband of 450 rpm, or as set in step 204 of a previous cycle of the routine. Then, in steps 142 and 144, upper limit and lower limit threshold values, THUL and THLL, respectively, are 20 derived from the rate limited speed control lever value, RSCLl Next, in step 146, minimum and maximum values 10 and 100, respectively, are established for both the THU. and THUL values. Finally, in step. 148, the THU and TH0L are established as minimum and maximum values, respectively, for the engine 25 speed threshold value, THRS, for use in calculating the engine speed error value EEBR in later step 150.

In step 150, an engine error (or droop) value, EEPP, is calculated from the statement EERB « (PSCL· - TBPS - ENGS) x 1.5, where PSCL is the rate limited speed control lever position 30 value from step 132, TBPS is the engine speed threshold value from steps 140 - 148 and 1.5 is an engine.error gain or sensitivity factor chosen so that the system is suitably sensitive to variations in the engine speed. The terms PSCL -THRS in step 150 can be viewed as an engine speed set point 35 below which the engine speed value, ENGS, must fall before a positive EEBR value will be generated. (Negative EERR values .are ignored due to later step 158.) This set point value will always be smaller than the RSCL value by the amount of the THRS value,' which, due to steps 140 - 148, can vary between 10 and 40 100 representing rpms of 100 and 1000, respectively. By 14 maintaining this set point always a certain amount less than the speed control lever position represented by the RSCL value, the implement is prevented from being continuously raised and lowered in response to the normal continuous and minor 5 fluctuations in the engine speed of the tractor. The minimum threshold value established in steps HO - 148 prevents too small a THRS value from being used in step ISO if the load command control 56 is adjusted when the speed control lever 14 is at reduced engine speed settings. The maximum threshold value established in steps 140 - 148 prevents too large a THRS value from being used in step 150 when the speed control lever 14 is moved to reduced engine speed settings after a stored threshold value, STHRS, has been set in step 204 when the speed control lever 14 is at large engine speed settings.

Then, in step 156, the EERR value is set equal to 0 if the FLAG value is equal to 1, as would be the case if it were set equal to 1 in step 196 of a previous cycle of the routine in response to a change in the LOOM value. Next, in a portion of the routine represented by step 158, lower and upper limits, 0 20 and 255, respectively, are established for the range of possible EERP values. As mentioned previously) step 158 prevents a negative EERR value from influencing the hitch and/or implement and prevents any overflow in the eight bit register containing the EERR value.

Next, in a portion of the routine represented by step 160, a rate limited engine speed error value, RERR, is derived from the engine speed error value', EERP. In short, when the EEPP value is increasing, then the RERR value is incrementally increased each cycle of the routine by a constant amount which can he no 39 greater than an amount corresponding to a rate of change of, for example, 420 rpm per’second. Alternatively, if the EERR value is decreasing, or if the EERR is reduced to zero at step 156, then the RERR value'is incrementally decreased each cycle of the routine by an amount corresponding to a rate of change of, for 35 example, 160 rpm per second. Thus, the RERR value will increase or decrease in a linear fashion as a function of time. The rate limited engine speed error value RERR is then used later in the routine in the determination of the load or combined error value, LERR. In this manner, the^sensitivity of the system to 40 changes in engine speed, that is, the rate at which the 15 implement is raised in response to a deteriorating or lugged-down engine speed and the rate at which the implement is lowered in response to a recovering engine speed, is limited to achieve a more stable control system by preventing over-shoot and ^ hunting. The limits are different under the deteriorating and recovering engine speed situations because it has been found to be desirable to permit a faster implement raising speed to prevent engine stalling, while the slower implement lower speed is desirable for system stability.

At step 170, the position feedback value, FDEK, is derived from the sensed position value A(l). Calibration factors are applied to the A(l) input to assure that the range of possible^, positions sensed by position sensor 60 will provide FDBK values with digital counts ranging from 0 to 255. The calibration 15 factors will vary, depending upon the type of position sensor and particular linkage structure between the position sensor and the implement. Then, in step 180, a filtered position value is determined according to the statement FFOS « (127 x FFOS + FDBKJ/128 so that the filtered position value FFOS is a weighted 70 average of the FDEK value from step 170 and the initial FPOS value of zero or the FFOS value set at step 180 of a previous cycle of the routine. The FDBK and FFOS values are employed later in the routine to determine a normalized position feedback term which becomes part of the combined load error value, LEBR, 25 while the FDBK value is used later to determine a position error value, FBFR.

Next, in step 182, a sensed draft force value DRAFT is determined by the statement DRAFT « (DRAFT + A(10))/2, so that the DRAFT value set in step 182 is an evenly weighted average of 30 the input A(10) from draft sensor 38 and the DRAFT value which is initially zero or is set in step 182 of a previous cycle of the routine. Next, in step 184, a filtered draft value FDPFT is determined by the statement FDPFT (127 x FDRFT + DPAFTJ/128 so that the FDRFT value set in step 184 is a weighted average of 35 the DRAFT value from step 182 and the initial FDRFT value of zero or the FDRFT value set at step 184 of a previous cycle of 'the routine. This DRAFT value is also employed later in the routine to determine the load or combined error value, LERR.

The FDRFT value is employed later in step 232 in determining a <0 DGAIN value. 1β In step 186, a scaled and zecoed load command value, LCOH, Is derived from the A(6) input of the load command sensor 58 so that LOOM values from C to 255 represent the Cull range of positions of the load command sensor 58.

In a portion of the routine represented by step 187, a rate- limited load command value, RLCOM, is derived from the LCOM value. In short, the RLCOM value is incremented or decremented each cycle of the routine in response to an increasing or a decreasing, respectively, in the LCOH value. The increment and 10 decrement are chosen so that for a maximum possible step change in the load LCCM value, the FLCCM value will attain the new LCOH value on the order of approximately 6 to 8 seconds.

In step 188, an engine error limit value, ELIH, is calculated according to the statement ELIM » (LCOH x .2) - 13, 15 where .2 and 13 are experimentally determined scaling and offset factors. The ELIM value is used later in steps 216 and 218, as described later in this description. Then, in step 190, lower and upper limits 10 and 255, respectively, are established for the ELIM value.

Next, in step 192, the LCCM value is compared to 250, the WVEL value is compared to 30 and the PCOM value is compared to 130. If the LCOM value is greater than 250, or if the WVEL is less than 30, or if the PCOM value is greater than 130, then the routine proceeds to steps 206 and 208 where the engine error, 25 -KERR, and slip error, SERR, values are both set equal to 0, after which the routine proceeds to step 212. Step 192 bypasses steps 194-204 to prevent recalculation of the set points in step 204 when a large draft force setting is chosen (LOOK>-250), when the vehicle is not moving (WVEL<30) and when the hitch and 30 implement are raised out of the ground (?CCM»130). Steps 192 and 206-208 also reduce the effect of wheel slip and engine droop under these same conditions. However, if none of these conditions are met, then the routine proceeds to step 194.

In step 194, the LCCM value is examined to determine if the 35 operator is changing-the setting of the load command knob 56, by comparing to zero the absolute magnitude of the difference between the rate-limited load command value, RLCOM, and the load command value, LCCM. if this difference is non-zero, then from step 194, the routine proceeds to step 196. In step 196, a 40 countdown counter or timer is initialized to a value, Gil, which 17 represents a minimum set point delay tiire period of, for example, 120 mlliseconds. Also, in step 196, the FLAG value is set equal to 1 and a stored filtered position value, or reference position value, SFPOS, is assigned the value of FPOS 5 from.step 180. After step 196, the routine proceeds to step 21? However, if (RLCOM - LCOH) equals zero in step 194, then the routine proceeds to step 198. Step 198 determines whether the counter set in step 196 has been counted down to zero. If the counter has not been counted down to zero, then the routine 10 proceeds to step 200, where the timer is decreased by 1 count, after which the routine proceeds to step 212. If the counter has counted down to zero in step 198, then the routine proceeds to step 202 where the value FLAG is compared to 1 and a load control mode value LCONT value is compared to 1. If, in step 15 202, both the FLAG and LCONT values are not equal to 1, then the routine proceeds to step 212. However, if both the FLAG and LCONT values are equal to 1, then the routine proceeds to step 204, where various set point values are calculated for use in other portions of the routine. For example, the slip set value, 2o SSET, is set equal to the sum of the filtered slip value, FSLP, (from step 114) plus 5. In this manner, wheel slip, which is less than 5% above the average slip value, FSLP, will not cause the implement to raise. The stored engine speed threshold value, STHBS, is determined by the statement STHRS “ RSCL - ELIM 25 + 10. Also, the stored filtered or reference position value, SFPOS, is assigned the filtered position value FPOS to provide . an operating reference for the draft control position feedback term in later step 228. Finally, in step 204, the changing load command flag value, FLAG, is set equal to 0 so that step 202 3q will prevent recalculation of the various set point values unless the FLAG value is reset to 1 at step 196 in a subsequent cycle of the routine. Thus, in response to a change in the signal from the load command transducer 58, the SEPR and PEPP values are momentarily modified or reduced towards 0 by 35 operation of steps 119 and steps 156 and 160, respectively.

Then after expiration of a certain variable delay period - following such a change in the load command signal, the set point values SSET and'STHRS are recalculated. The recalculated set point values are maintained until they are again recalculted 40 in response to another change in the load command signal. 18 This variable delay period will be the minimum delay period represented by the Gil value when the load command changes are small. However, larger changes in the LCOM value will require a longer time for the PLCOM value to 'catch up* to the new LCOH 5 value, on the order of a few seconds, due to step 167. Thus, step 194 will prevent step 204 from recalculating the set points until this longer delay time of a few seconds has expired. Furthermore, due to step 187, this longer delay period will be proportional to the amount of change in the LCOH value. In 10 this manner, steps 194 - 204 operate to momentarily reduce the engine speed and wheel slip error values' towards zero whenever the operator makes a predetermined change in the load command control 56. This is done so that the draft control operating point can be set in steps 226, 228 and 234 without interference IS from the wheel slip and engine speed error inputs. This assures that the draft force responsive aspect of this control system is normally the primary or dominant control aspect.

Also, when a change occurs in the load command value, LOOM, the reference position value, SFPOS, will be repeatedly 20 redetermined or set equal to the current filtered position value, FPOS, in steps 196 or 204, until the rate limited loed command value, RLCOM, re-attains the value of LCCM, whereupon the reference position value, SFF0S, will remain unchanged for use as a reference value for the normalized position feedback 2S term (FDEK - SFPOS) for use in step 228 until new changes in the LOOM value occur. Since the FPOS value will normally be only -slightly different from the FDEK value, it follows that this normalized position feedback term (FDBK - SFPOS) will have a modified or reduced value and effect in step 228 as long as the 30 RLCOM value is "catching up" to the new LCOM value as a result of a change in the LCOM value. Furthermore, since the SFPOS value follows the smoothly changing FPOS value, abrupt changes or discontinuities in the LERP value in step 228 are prevented,, thus resulting in a smooth implement control characteristic. · 35 After step 204, the routine proceeds to step 212 where a downshift light value, DLIT, is set to 0.

Hext, at step 214, the wheel velocity value, WVEL, is compared to .a value 30'which represents, for example, a .3 meters/sec. tractor velocity. If WVEL is less than 30, the 40 routine proceeds directly to step 224. However, if WVEL is not 19 less than 30, then the routine proceeds to step 216 where the rate limited engine speed error value RERR is compared to the ELIM value. If the RERR value is not greater than the ELIH value, then the routine proceeds directly to step 224. However, 5 if RERR is greater than ELIH, then the routine proceeds to step 218, where the RERR value is assigned the value of ELIM, thereby limiting the engine error value, RERR, to the ELIM value, which, as set in step 188, is proportional to the setting of the load command knob, as represented by the LCCM value. After that, the 10 routine proceeds to step 220.

Thus, in the situation where the tractor and implement start to go up a hill, the implement draft force and the resulting DRAFT value from step 182 may decrease due to the decreased velocity of the tractor and implement. How, in ordinary draft 15 control systems, a reduction in sensed draft force tends to cause a lowering of the implement. However, with the present system, as the engine speed decreases, the resulting engine error value, RERR, will increase. This increased RERR value will increase the combined load error value LEPP in step 228, 20 thus countering the effect of the reduced draft force caused by the reduced ground speed. If the tractor continues to lose ground speed because of the severity of the hill, and the RERR value becomes equal to ELIM, then RERR will be set equal to ELIM, and the hitch will be prevented from further raising the 25 implement. In this situation, it would be preferable for the Operator to increase the engine speed, and thus reduce the RERR value by downshifting the tractor transmission to a lower gear. Accordingly, the routine limits the amount by which the implement is raised in response to reduced engine speed by 20 limiting the RERR value to the value ELIM in step 218.

Furthermore, with steps 220 and 222, which are optional, the routine generates a signal (DLIT * 1) when the slip error value SERR is less than the engine error value PEER and when LCONT equals 1, as set in step 268, when the implement is being 35 controlled in response to the combined load error value, LERP.

A signal representing the DLIT value may be communicated to an indicator device (not shown) in the tractor cab to provide the operator with a signal when it is appropriate to downshift the transmission to a lower gear. As an added feature, step 214 20 prevents generation of this warning signal when the tractor is not moving. Without steps.220 and 222, the routine would proceed from step 218 to step 224.

At step 224, an activity gain value, GAIN, is calculated 5 from the equation GAIN « A(4) x .004 + .5, where A(4) is a value representing the signal from the activity control transducer 72. Since the A(4) value may be varied from 0 to 2SS, the factors .004 and .5 are applied to A(4) to obtain GAIN values in a desired range of .5 to 1.5. In step 226, an interim coabined 10 error value, COMH, is determined by the equation COMM.· tCOH -SERR - PEPR. Next, at step 228, an interim combined load error value,’ LERR, is calculated from the equation LERR * (-COMM + DPAFT - (FDEK .- SFPOS) x .4) x GAIN. The normalized position'" feedback term {FDBK - SFPOS) helps to prevent overshooting and rough hitch and- implement movements. The (.4) value is a constant sensitivity value which is optimized for good depth control during tractor pitching in unevenly contoured, terrain and during operation in soil of non-uniform density. As described previously, the SFPOS value is redetermined to 2o establish a new reference value for the normalized position feedback term in response to changes in the load command, LOOM, value.

In step 230, if the rate limited position command value, RCOH, is greater than a value, for example, 130, then a dynamic 25 feedback gain value, DGAIN, is set equal to 0. The PCOM value is initially 0 from step 99 or is determined in steps 252 or *254 of a previous cycle of the routine. Then, in step 232, if the filtered draft value, FDRFT, from step 184, is greater than 85* . of the load command value, LOOM, then DGAIN„is set equal to 1.

Then, at step 234, the DERR value is recalculated according to the equation LERR « (LERR - (FDBK - FPOS) X DGAIN) X 4, where DGAIN is a dynamic feedback gain factor which is applied to the negative feedback factor, (FDEK - FPCS), and where 4 is an overall loop gain factor. The overall loop gain factor may vary 35 over a range of values, for example, 3-5, depending upon.the particular hitch, implement and vehicle that are involved.

Thus, the ultimate combined load error value, LERR, in step 234 is an arithmetic function or combination of the load command value, LCOM, from step 186, the draft force value, DPAFT, from 40 step 182, the rate limited engine speed error, RERR, from steps ai 160 or 218, the wheel slip error value SERF from step 126, the implement position value, FDBK, from step 170, the filtered position feedback value, FPCS, from step 180, and the GAIN value from step 224. The position feedback adds stability to the 5 system operation. By changing the GAIK value by manipulating the activity control 70 and its corresponding transducer 72, the overall sensitivity or activity of the control system may be adjusted, at least with respect to the component values of the ItERR value. In step 234, the term (EDEK - FPOS) provides a 10 negative feedback proportional to the ratd of change of the implement position which adds further stability to the system by, preventing overshooting of the implement when rapid changes in draft force occur.

Step 230 operates to reduce the effect of the dynamic 15 feedback term in step 234 when the implement is raised out of the ground so that rapid implement movement can be obtained.

Step 232 operates to increase the effect of the dynamic feedback term in step 234 when the implement is in the ground and the average sensed draft force represented by FDRFT achieves 85% of 20 the desired draft force level represented by the LCCH value.

Note that regardless of the cause, a reduced tractor velocity can cause a decrease in the draft force and a corresponding decrease in the- DRAFT value set at step 182. The control system tends to respond to a decreased DRAFT value by 25 lowering the implement, just as in a conventional pure draft responsive system. Dependirig on the initial cause, this implement lowering can further reduce the ground speed of the tractor, thus reducing the engine speed, and possibly stalling . the engine. However, this implement lowering response is 30 counteracted in the present control system in that reduced travel velocity also causes an increase in the filtered rate limited engine speed error value, PEPI>, which, through steps 226 and 228, offsets the decrease in the DRAFT value, thus counteracting the draft force responsive implement lowering and 35 helping to prevent engine stalling.

Also note that the combined or load error value, LERR, will increase if the wheel slippage increases, as represented by an Increasing SERR value in step 226. This slippage induced increase in the LERR value tends to raise the implement, thus 4o shifting weight to the rear wheels, increasing the traction and 22 reducing the slippage thereof. Onlike the situation with respect to engine speed, wherein the* engine speed induced implement raising is limited by the operation of steps 188, 190, 216 end 218, the amount by which this control system will raise 5 the implement in response to wheel slip is not limited. Xn otherwords, to prevent the tractor from getting stuck, the control system will raise the implement entirely out of the ground.

Thus, the present control system includes two slip-related 10 aspects, the first related to weight transfer and the second related to preventing the tractor from becoming stuck. Also, in vet or low traction soil, this control system will establish a, maximum implement depth.

Next, in a portion of the routine represented by step 23S, 15 upper and lower limits 255 and -254, respectively, are established for the LERR value to prevent register overflow. Then, at step 238, the position command value, ROOM, is derived from the value of A(5) which is the output of the operator controlled position command sensor or transducer 54 and which 20 represents a desired position of the implement. Next, at step 240, a lower position limit value, LLIM, is calculated from the equation LLIM (A(2) x .6) + 3, A{2) representing the output of the operator-controlled lower limit transducer 53 which may be positioned at a suitable location in the operator's 25 'compartment. The factors .6 and 3 are chosen so that the lower limit value LLIM may be varied over the lower 60t of the range of positions which the PCOH value represents. Next, at step 242, an upper position limit value, ULXM, is calculated from the equation OLIM » (A(3} x .3) + 178, A(3) representing the output 30 of the operator-controlled upper position limit transducer 51 also located in the operator’s compartment. The factors .3 and 178 are chosen so that the upper limit value OLIH may be varied over the upper 30t of the range of positions represented by the PCOM value. Next,' in a portion of the routine represented by 35 step 244, the LLIM and OLIM values are established as the minimum and maximum values between which the PCOH value may range. In this manner, the position command value, PCOM, is limited to between the lower and upper limit values LLIM and OLIM, regardless of the position of the position command lever 23 52 and its transducer 54. Of course, an alternative ,to this would be to physically limit the position range of the position lever 52 with mechanical stops.

Next, in step 246, the status of an operator-controlled 5 implement lift/lower rate selecting switch 74 is examined. If the switch 74 is closed, as when the operator desires a rapid Implement movement rate, then a FASTSL· flag value is set equal to 1. If the switch 74 is open, as when the operator desires a slow implement movement rate, then the FASTSL· flag value is set 10 equal to 0.

Then, step 248 operates to bypass the rate limiter steps 250 - 254 and proceed directly to step 256 if the system is in its. load control mode (as represented when ICONT * 1 in step 268) or if the PCOM value is less than the FDEK value, as would be the 15 case where the implement is resting on the ground. However, if these conditions are not met, then the routine proceeds directly to step 250 where the routine is routed to steps 252 or 254, depending upon whether the FASTSL· value equals 1 or 0, respectively. In the portion of the routine represented by step 20 252, the PCOM value is examined to determine whether it has increased or decreased from its value in a previous cycle of the routine. If it has Increased, then a rate limited position command value, fiCGM, is incrementally increased during each cycle of the routine by a set amount, 1.2, which corresponds to 25 a rate of implement movement which would result in the implement being moved through its full range of travel in a certain time period, for example, 2 seconds. Similarly, the PCOM value is decrimented if the PCOM value has decreased. The incrementing · or decrimenting of the PCOM value continues until the PCOM value 30 attains the new PCOM value. In the portion of the routine represented by step 254, the PCOM value is similarly incremented or decrimented, except that it is incrementally changed by an . ammount, .7, which corresponds to a rate of implement movement which would result in the implement being moved through its full 35 range of travel in a'longer time period, for exarple, 4 seconds. Thus, steps 250 - 254 provide for a variable rate of implement lifting or lowering which is independent of the amount of weight or type of implement moved by the system within the limitations of the hydraulic system. After steps 252 or 254, ’ 40 the routine proceeds to step 256. 24 At step 256, the position error value, PEPP, is calculated from the equation PERR « (PCOH - FD9K) x 7, where the BCOH value is set by previous steps 238, 244, 252 and 254, where the POPE value is set by step 170 and where 7 is a position gain value 5 which is chosen so that the system has a satisfactory sensitivity to the difference between the desired and sensed implement positions. Then, at a portion of the routine represented by step 258, upper and lower limits +255 and -255, respectively, are established for the range of PEPP values to 10 prevent overflow problems in the register containing the PEPP value.

Next, in step 260, the PEPP and LERP values are compared.— If the PERR value is larger than the LERR value, then the valve command value, EVCO, is set equal to the PEPP value in step 15 262. On the other hand, if the LERR value is greater than or equal to the PEPP value, then HVCO is set equal to the LERR value in step 264. Thus, depending upon which of the values PERR and LEER is most positive, the control system operates to raise or lower the implement in response to one or the other of 20 the position error signals, PERF, cr the load or combined error signal, LERR.

The digital HVCO value from steps 262 or 264 is converted to a pulse width modulated signal via PIT 320, for example, and is applied to the torque motor of'the control value 42 so that an 25 increase in the HVCO value will tend to raise the implement (not shown). The speed of implement movement will be proportional to the magnitude of the HVCO value. Raising the implement reduces the LERR or PERR value and reduces the HVCO value until a steady state condition is obtained wherein the HVCO value is 30 minimized. Conversely, a lowering of the BVCO value causes a lowering of the implement until a new steady state condition is achieved.

After steps 262 or 264, the routine proceeds to step 266, where a position differential value, PERR-PHXS, is compared to a 35 load differential value, LERR-LBYS, where LERR is the combined load error value from step 234, where LHYS is a load hysteresis value which is initially 0 in step 9S or is set equal to 0 or 10 in steps 268 or 270, respectively, where PEPR is the position error value from step 256, and where PHYS is a position 40 hysteresis value set initially at zero in step 99 or is set 25 equal to 10 or 0 in steps 268 or 270f respectively. If, in step 266, the position differential value'is less than or equal to the load differential value, (as would be the case when the LEPR value is the dominating error value in steps 260 and 264), then 5 the routine proceeds to step 268 where the flag value LCONT is set equal to 1, the load light value LLIT is set equal to 1, the position light value PLIT is set equal to 0, the position hysteresis value PHYS is set egual to 10 and the load hysteresis value is set equal to 0. The LCONT value is set.egual to 1 so 10 that it is possible to activate the downshift light through the operation of steps 226 and 228. Thp load light value LLIT is 6et equal to 1 so a signal representing the LLIT value may be communicated to a load control mode indicator device 80, located in an observable position in the tractor cab, to inform 15 the vehicle operator that the implement is being controlled in response to the combined load error value, LEPR. Conversely, the PLIT value is set equal to 0 so that a position control mode indicator device 82, also located in an observable position in the tractor cab, may be disabled in this situation. The 20 position hysteresis and load hysteresis values, PHYS and LHYS, are set equal to 10 and 0, respectively, so that during subsequent cycles of the routine, step 266 will continue to direct the routine to step 268 until the control system switches to a position error 'dominated control mode due to a significant 25 increase in the PEPR value relative to the LERP value. In this manner, a possibly annoying intermittent or flickering operation of the load and position control mode indicator devices 80 and 82 may be prevented.

If, however, in step 366, the position differential value is 30 greater than the load differential value, (as would be the case when the PERR value is the dominating error value in steps 260 and 262), then the routine proceeds to step 270, where the flag value, LCONT, is set egual to 0, the LLIT value is set ecual to 0, the PLIT value-is set equal to 1, and PHYS value is set equal 35 to 0, and the LHYS value is set egual to 10. In this case, the LCONT value is set egual to 0 to prevent the actuation of the • downshift indicator device through the operation of steps 220 and 222. The LLIT and PLIT values are set egual to 0 and 1, respectively, so .that the corresponding load and position 40 control indicators 80 and 82 may be disabled and energized. 26 respectively, to give an indication that the system is operating in a position error responsive mode.' The hysteresis values, PHYS and LHYS are set egual to 0 and 10, respectively, to prevent intermittent operation of the load and position control 5 mode indicators 80 and 82, in a manner similar to that previously described with reference to step 268. After steps 268 or 270, the routine returns to step 100 via step 272. The entire routine may be repeated at a 100 Bz rate, although other repetition rates may also be suitable.

Cue to step 260, where one or the other of the I.EFR or PERP values is selected, this control system is essentially a split control system, rather than a mixed control system where the control signal is a mixture of the position error, PEPF, and combined load error, LEP.R. As a result, a steady state or 15 equilibrium condition may be attained wherein only one of the LERR or PEER values is maintained at zero, while the other value could be negative. In this case, the negative error value is ignored by the most positive select function of steps 260 -264. An example of this situation is where the PEPP value is 0 20 because the position command, PCOH, and position feedback, FDBK, values are egual, but where the load error, LERR, value is negative because the load command, LCOM, value represents a larger draft force than the actual sensed draft force represented by the DRAFT value. In order to lower the implement 25 to operate at an increased amount of draft force, the operator must reduce the position command value, PCOH, so that the PERF value in step 256 becomes at least as negative as the LEFF value, thus permitting steps 260 - 264 to select less negative of the LEFF and PEER values to thereby lower the implement and 30 increase the draft force operating point. The FLIT value generated by step 270 can provide the operator with an ' indication of when such a situation occurs by informing him that the control system is in the 'position error mode and thus will not respond to an-increased LCCM value unless the PCOH value is 35 also adjusted.

An analogous situation occurs when the LERE value is 0 and • the operator desires to lower the implement by reducing the PCOH value. Except in this case, the load light value, EDIT, is set egual to 1 by step 268 to inform the operator that the control 40 system is in its load control mode and that the LCOM value must 27 be increased to permit lowering of the implement through lowering of the PCOM value.

Note, however, that the hitch will be automatically raised whenever either of the LERR or PERR values is increased to a 5 value which is more positive than the other. This permits the control system to automatically raise the implement to reduce wheel slippage and engine stalling, as desired.

The conversion of the above flow chart into a standard language for implementing the algorithm described by the flow 10 chart in a digital data processor, such as the microprocessor 310, will be evident to those with ordinary skill in the art.

It should also be noted that the particular numerical values given in the preceeding description are merely exemplary and could be changed for many reasons, including tailoring the 15 system for a different implement or vehicle, without departing from the scope of the invention. It would also be within the scope of this invention to control a towed implement with the previously described drawbar draft sensing, or with a T-bar on the draft links and draft link sensing, with position sensing on 20 a remote cylinder of the towed implement and with another electrically-operated 4-way, 3-position control valve for controlling fluid communication to the- remote.cylinder or with a diverter valve for diverting fluid from a solenoid-operated rockshaft control valve to the remote cylinder on the towed 25 implement. Similarly, it is within the scope of this invention to control a semi-integral implement with one or more remote cylinders in series with the rockshaft cylinder with draft sensing in the hitch draft linkages and with position sensing in.' the rockshaft or in the rockshaft cylinder.

While the invention has been described in conjuction with a specific embodiment, It is to be understood that many alternatives, modifications, and variations will be apparent to those skilled in the art in light of the aforegoing description. Accordingly, this invention is intended to embrace 35 all such alternatives, modifications, and variations which fall within the spirit and scope of the appended claims. i ¥ 28 It will be noted that the slip error signal SERR is the difference between the SLIP parameter signal and the slip set-point signal SSET and that SSET is derived from the SLIP parameter signal itself by way of block 114 where FSLP is derived from SLIP 5 and block 204 where SSET is derived from FSLP. STHRS and SFPOS are other examples of set point signals which are derived (block 204 again) from the corresponding parameter signals.

Claims (2)

29

1. A working depth control system for a vehicle having connecting means for attaching a ground-penetrating Implement thereto and an actuator means for raising and lowering the implement to vary the ground penetration thereof in response to a control signal 10 15 derived from a combined signal formed from a plurality of error signals representing differences between actual parameter signals such as draft force, driven wheel slippage and tractor engine speed, said machine system derived from a machine system comprising the vehicle, the attaching means and the implement, and corresponding set-point signals, the combined signal being formed as a combination of at least three parameter error signals, each of which is the difference between a sensed parameter value and a corresponding set-point value, the sensed parameter values being engine speed, wheel slip and draft force and the engine speed set-point value and draft force set-point value being provided by operator controls .

2. A working depth control system according to Claim 1, substantially as hereinbefore described with particular reference to and as illustrated in the accompanying drawings. Dated this the 16th day of August, 1988. P. R. KELLY & CO. EXECUTIVE 27 Clyde Road, Balisbridge, Dublin 4 AGENTS FOR THE APPLICANTS.