RU2703086C2 - Autonomous loading vehicle controller - Google Patents

Abstract

FIELD: physics.

SUBSTANCE: embodiments of implementation may contain at least one controller, which controls the bucket and/or autonomous loading-and-transport vehicle in accordance with at least one sensor signal, at that this at least one sensor signal characterizes interaction between scoop and rock bulk during scooping. Some embodiments of implementation include` at least one compliance adjustment device and, if required, at least one iterative learning controller (ILC), which uses feedback from at least one previous digging for modifying said at least one sensor signal output to said at least one controller.

EFFECT: invention relates to device for control of digging and method of control of digging for autonomous loading and transport vehicle (ALV), used in such applications as development of deposits, construction and exploration works.

25 cl, 20 dwg, 2 tbl

Description

This application claims the benefits of the filing date of US patent application No. 61 / 931,243, filed January 24, 2014, and patent application US No. 62 / 033,904, filed August 6, 2014, the contents of which are fully incorporated herein by reference.

FIELD OF TECHNOLOGY

This invention relates to the management of earth moving / material handling vehicles. In particular, this invention relates to autonomous or semi-autonomous driving of earth moving / loading vehicles.

BACKGROUND OF THE INVENTION

Autonomous (or robotic) earth moving / loading vehicles are of interest in the mining and construction industries, where the task is to remove operators from hazardous environments, increase the degree of utilization and productivity of machines, and reduce operating costs. Autonomous excavation is also of interest in the study of the moon or planets, on which excavation cannot be easily performed by remote control.

In the mining and construction industries, autonomous excavation typically involves the development of crushed rock using a material handling machine (LHD). What makes robotic excavation a difficult task is the nature of the interactions between the bucket and the rock. Operational characteristics are strongly influenced by the conditions of interaction between the machine and the environment surrounding it. For example, the resistance that the bucket encounters when it tries to penetrate into the rock bulk can vary significantly depending on the properties of the medium (e.g. density and hardness), the geometry of the rock bulk, and the distribution of sizes and shapes of the rock particles. Indeed, it will be very difficult to determine in advance the exact nature of future bucket – rock interactions before performing any particular mining / excavation operation.

Previous designs have attempted to automate excavation / excavation by using a control device and forces measured on a bucket to offset between separate scooping trajectories for excavation from a rock pile or using a malleability control device to scoop specified soils. Both of these control / regulation devices were tested only in relatively homogeneous materials, and they did not work properly in a collision with underground bumps. Consequently, they were poorly suited for the autonomous development of typical rock bulk in real situations in which they often encounter underground obstacles. Other developments that used a control device based on fuzzy logic gave unstable results and are difficult to implement and maintain as a commercial product. A control device has been proposed based on compliance control and using measured forces to control the speed of the bucket actuator, but it has never been implemented or tested.

SUMMARY OF THE INVENTION

This document describes a scoop control device for an autonomous or semi-autonomous truck (ALV), comprising: at least one controller that controls a bucket and / or autonomous truck in accordance with at least one a sensor signal, while this at least one sensor signal characterizes the interaction between the bucket and bulk rock during scooping; however, this at least one sensor signal is received from at least one sensor connected to one or more actuators other than the bucket actuator, or to one or more drive elements.

In one embodiment, the scoop control device further comprises at least one iterative learning controller (ILC), which uses feedback from at least one previous scooping to modify a given at least one sensor signal issued to a given at least one controller.

This document also describes a scoop control device for a stand-alone truck, comprising: at least one controller that controls a bucket and / or stand-alone truck in accordance with at least one sensor signal, when this given at least one sensor signal characterizes the interaction between the bucket and the bulk of the rock during digging, and at least one iterative learning controller that uses feedback t, at least one previous digging for modification of at least one sensor signal outputted present, at least one controller.

In one embodiment, the at least one controller comprises at least one compliance control device (admittance controller). In one embodiment, the at least one sensor signal is obtained by measuring the force perceived by the boom actuator. In another embodiment, the at least one sensor signal is obtained by measuring the force perceived by the drive element.

In one embodiment, the malleability control device may adjust the bucket speed. At least one compliance control device may comprise an adaptive compliance control device. The adaptive compliance control device can dynamically adjust at least one parameter in response to the difference between the sensor signal and the predetermined signal.

In one embodiment, the at least one controller converts one or more sensor signals into a range of possible bucket speeds or speeds of the autonomous truck by using at least one of proportional, integral, and differential control.

In one embodiment, the at least one controller can convert the total force deviation into a range of possible measured forces by using at least one of proportional, integral, or differential control. In another embodiment, the at least one iterative learning controller can convert the signal from previous scooping into changes in the output of the scoop control device by using at least one of proportional, integral, and differential control.

In one embodiment, the scoop control device can modify a given at least one sensor signal issued to the control device so that the bucket speed or the speed of the autonomous truck changes.

In one embodiment, the scoop control device may comprise at least one position adjuster that controls at least one of the movement of the bucket of the autonomous truck to at least one selected position and the movements of the autonomous truck relative to the bulk of the rock. In another embodiment, the scoop control device may comprise a first iterative learning speed controller that changes the speed of the autonomous truck based on a sensor signal characterizing the interaction between the bucket and the bulk of the rock during at least one previous scooping, and a second iterative learner a regulator that modifies the sensor signal obtained from measuring the deviation of the force acting on the bucket and boom for at least one p edyduschem digging.

In yet another embodiment, the scooping control device may comprise a first iterative learning controller that modifies the sensor signal outputted by the malleability control device, and a second iterative learning controller that modifies the sensor signal outputted by the malleability adjustment device, the modification being based on feedback from at least one previous scooping.

This document also describes a software tool for use with a scoop control device for a self-contained material handling vehicle and comprising a computer, the software tool comprising: a computer program stored on a non-volatile storage medium compatible with a computer, and a computer the program contains commands for controlling a computer to perform one or more of: receiving at least one sensor signal from, in m at least one sensor interacting with one or more actuators other than the actuator of the bucket, or one or more drive elements, while this at least one sensor signal characterizes the interaction between the bucket and the bulk of the rock during scooping, and control the bucket and / or autonomous truck in accordance with this at least one sensor signal.

This document also describes a software tool for use with a scoop control device for a self-contained material handling vehicle and comprising a computer, the software tool comprising: a computer program stored on a non-volatile storage medium compatible with a computer, and a computer the program contains commands for controlling a computer to perform one or more of: controlling a bucket and / or autonomous material handling vehicle means in accordance with at least one sensor signal, while this at least one sensor signal characterizes the interaction between the bucket and the bulk of the rock during scooping, or instructing the iterative learning controller to use feedback from at least , one previous scooping to modify this at least one sensor signal, while modifying this at least one sensor signal changes the control of the bucket and / or autonomous loading trans tailor tool.

This document also describes a method for controlling an autonomous loading vehicle, including: receiving at least one sensor signal from at least one sensor interacting with one or more actuators other than the bucket actuator, or one or more driven elements, while this at least one sensor signal characterizes the interaction between the bucket and the bulk of the rock during digging; and controlling a bucket and / or autonomous truck in accordance with the at least one sensor signal.

This document also describes a method for driving an autonomous truck, including: controlling a bucket and / or autonomous truck in accordance with at least one sensor signal, while this at least one sensor signal characterizes the interaction between the bucket and the bulk of the rock during digging; and modifying the at least one sensor signal by using at least one iterative learning controller, which includes feedback from at least one previous scooping, while modifying this at least one sensor signal changes the control of the bucket and / or autonomous truck.

In one embodiment, the method may include modifying the at least one sensor signal output to the control device / controller such that the bucket speed or the speed of the autonomous truck changes. The control may further include dynamically adjusting at least one parameter in response to a difference between the sensor signal and the predetermined signal.

The method may further include controlling at least one of the movement of the bucket of the autonomous truck to at least one selected position and the movement of the autonomous truck relative to the bulk of the rock. The method may further include varying the speed of the autonomous truck based on a sensor signal characterizing the interaction between the bucket and bulk in the course of at least one previous scooping, and modifying the sensor signal obtained from the measurement of the deviation of the force acting on the bucket and an arrow with at least one previous scooping. The method may further include modifying the sensor signal outputted to the malleability control device, and modifying the sensor signal outputted by the malleability adjustment device, the modification being based on feedback from at least one previous scooping.

In one embodiment, the control may include converting the total force deviation into a range of possible measured forces by using at least one of proportional, integral, and differential control. The method may further include converting the signal from the previous scooping into changes in the output of the scoop control device by using at least one of proportional, integral, and differential control.

In another embodiment, the payload can be controlled based on a tear-off state parameter or by modifying such a parameter.

BRIEF DESCRIPTION OF THE DRAWINGS

For a better understanding of the invention and for a clearer demonstration of how it can be implemented, embodiments will be described by way of example with reference to the accompanying drawings, in which:

FIG. 1A is a schematic illustration of a stand-alone truck;

FIG. 1B-1D are schematic views of a stand-alone truck during three phases of scooping, corresponding to plunge (FIG. 1B), scooping (FIG. 1C) and separation (FIG. 1D);

FIG. 2A and 2B are block diagrams of generalized embodiments of a scoop control device;

FIG. 3A is a block diagram of a scoop control device in accordance with one embodiment;

FIG. 3B is a block diagram of an example scoop control logic used in embodiments of a scoop control device;

FIG. 3C is a block diagram of a scoop control device in accordance with one embodiment, which includes a malleability control device;

FIG. 3D is a block diagram of a scoop control device in accordance with one embodiment, which includes an iterative learning controller (ILC);

FIG. 3E is a block diagram of a scoop control device in accordance with another embodiment that includes an iterative learning controller;

FIG. 4A and 4B are flowcharts showing the generation of force corrections for the boom and bucket in accordance with embodiments;

FIG. 4C and 4D are flowcharts showing the generation of adjustments to the position of the throttle infeed regulator in accordance with embodiments;

FIG. 5 is a graph showing scoop performance values for 57 scoop attempts using the experimental setup of Example 1;

FIG. 6A and 6B are graphs showing predetermined profiles of forces acting on the boom and bucket, respectively, and predetermined forces (in rectangular frames) used by the malleability control devices of the boom and bucket in the embodiment described in Example 2;

FIG. 7 is a graph showing a predetermined and actual rate of change of a plunging force acting on an arrow in accordance with the embodiment described in Example 2; and

FIG. 8 is a graph showing predetermined forces acting on the boom and bucket and used to calculate the total deviation for each scooping attempt (dark shading for negative deflection; light shading for positive deflection) for the embodiment described in Example 2.

DETAILED DESCRIPTION OF EMBODIMENTS

As used herein, the term “self-propelled truck” (ALV) is intended to mean a substantially autonomous, semi-autonomous or robotic excavator or material handling truck (LHD) used in accordance with the embodiments described herein.

As used herein, the term “actuator” is intended to mean a component of a self-contained material handling vehicle that causes a change in configuration and / or movement of the vehicle. The actuator may perform a function based on a command from a control device / controller / controller. For example, a vehicle configuration may include the position and / or orientation of a boom or scoop tool and / or the position and / or orientation of a stand-alone truck.

As used herein, the term “drive element” is intended to mean a component of a self-contained material handling vehicle that is actuated by an actuator, such as an arrow or a scoop tool, or an actuator that is not currently receiving a command, but exposed by another actuator.

As used herein, the term “bucket” is intended to mean a generally intended tool for scooping a stand-alone material handling vehicle, which may comprise a bucket, knife, chisel, pitchfork, probe, cutting tooth or other tool known in the art.

As used herein, the term “bulk rock” is intended to mean generally the material loaded by an autonomous material handling vehicle. It should be understood that the material may be any type of material or composition that may be associated with earthmoving, construction, mining and exploration, such as, for example, earth, sand, gravel, ore, slag, salt, rock fragments, regolith or any combination thereof, but possible materials are not limited to the above.

In the sense used in this document, the term "scooping" is intended to refer generally to the actions carried out by an autonomous truck to perform a given function by using its bucket. For example, a given function may consist in filling the bucket with material from the rock pile, and the term “dig” can be considered equivalent to the term “excavate”. However, other actions (eg, “loading”) can also be performed and may, if required, involve the use of other scooping tools. The scooping control, which is performed by the autonomous material handling vehicle, is carried out by means of the scoop control device embodiments described herein.

As used herein, the term “modify” means “change”, “adjust” or “change” a quantity or value, for example, increase or decrease a quantity or value. The value or value may relate to the sensor signal. The modification may be performed in accordance with a mathematical operation or function and / or may be performed in accordance with a constant.

Embodiments of the scoop control device for autonomous material handling vehicles described herein provide efficient autonomous excavation / digging for a wide range of materials in applications such as field development, construction, and exploration. Embodiments are particularly effective for bulk rock, including crushed rock with arbitrary sizes of rock fragments that may be encountered, for example, in applications such as mining and construction.

The generalized autonomous material handling vehicle is shown in FIG. 1A. As shown in FIG. 1A, a self-contained material handling vehicle includes a bucket 1 attached to the boom 2. The bucket moves by actuating the linear actuator 3 (tilt) of the bucket, while the boom moves by actuating the linear actuator 4 ( lifting) arrows. These actuators, which may be electric, hydraulic, pneumatic, or a combination thereof, may be equipped with linear sensors or code angles to determine the configuration and / or movement of the bucket. Each actuator has a cylinder side and a rod side, shown respectively with reference numbers 7 and 8 for the bucket actuator 3. The boom and bucket actuators, respectively designated 4, 3, are connected to the vehicle 5, which can ensure the movement of the boom and actuators to a predetermined location within the working space. During loading, the vehicle moves the boom and bucket forward into the rock bulk 6 (for example, in FIG. 1B). The interaction between the bucket and the bulk of the rock (for example, Fig. 1C) causes changes in pressure both on side 7 of the cylinder and on side 8 of the rod of both linear actuators (for example, Fig. 1D) until the bucket is removed bulk rock.

In all block diagrams of FIG. 2A, 2B, 3A-3E and 4A-4D, signal descriptions are shown in blocks with dashed lines. As shown in the block diagram of FIG. 2A, a standalone material handling vehicle 10 typically interacts with a bulk rock 6. The sensors generate 14 sensor signals characterizing the interaction between the bucket and the bulk of the rock (for example, reaction forces 40), and signals characterizing the movement of one or more bucket actuators 50. Sensor signals can be generated using one or more sensors or a combination of sensors selected from an accelerometer, a force sensor, a pressure sensor, a torque sensor, a torque sensor and a strain gauge, but the possible sensors are not limited to the above. The bucket speed can be measured by using one or more sensors or a combination of sensors, measuring transducers and the like, selected from an accelerometer, an adjustable differential transformer with a linear characteristic, a device for measuring wave reflection (e.g., sonar, laser sensor, infrared sensor, video encoder, optical encoder) and potentiometer (e.g. string, linear or angular), but possible sensors and transmitters are not limited to sheukazannymi. The sensor signals are used by the scoop control device 20 together with parameters 16, such as predetermined forces 12, to generate control signals 18 that control the autonomous material handling vehicle.

Embodiments of the scoop control device may include or use a sensor system 30 and controllers / control devices for controlling the scooping mode performed by the autonomous material handling vehicle. Further details are shown in the block diagram of the generalized embodiment of FIG. 2B. The sensor system 30 includes at least one sensor 32 and, optionally, a signal driver 34 that provides the sensor signal as an input signal to the scoop control device 20, which may include a logic device 22 and a memory 24. There may also be manual controls 26 and an operator interface / console 28 are provided. One or more sensors may interact with the drive element 64 of the autonomous truck. For example, the controllers may include an actuator control device 60 for moving the boom and bucket actuators 62 and the plunge position, for introducing the autonomous loading vehicle into the rock bulk, and controlling the movement of the autonomous loading vehicle forward throughout the entire scooping. The sensor system may detect that a force threshold has been reached (e.g., 40 in FIG. 2A), after which the scoop control device 20 can use the flexibility control devices in the actuator control device 60 to control the speed of the boom and / or bucket actuators 62 depending on the measured forces. The sensor system may detect that the shaft of the bucket actuator is fully extended (for example, 50 in FIG. 2A), after which the autonomous truck can be moved forward and the position adjuster can be used to raise the boom to the weighing position. The sensor system 30 may include at least one linear or angular sensor for each actuator (e.g., boom and bucket) and at least one force sensor for each actuator. In one embodiment, the force sensors include one or more pressure sensors on each actuator (for example, one on the cylinder side and one on the rod side of the hydraulic actuators). If required, the sensor system may include a sensor for measuring forward movement of the autonomous truck. For example, a sensor may include one or more of a code goniometer on the wheel, an inertial sensor for detecting initial contact with the bulk of the rock, and a vision system for detecting and / or evaluating and / or determining characteristics of the state of the surface of the bulk of the rock. The vision system may include a distance measuring system / ranging system capable of generating a three-dimensional image of the surface of the bulk rock. A three-dimensional image can be used to select the point of contact between the bucket and the bulk of the rock so that the time and effort when scooping are minimized. In these embodiments, the control device may include a proportional, integral, or differential controller, or any combination thereof.

Embodiments of the scoop control device are shown in FIG. 3A-3E. The scoop control device may include one or more flexibility control devices 20A (FIG. 3A). Compliance control devices respond to changes in strength by changes in speed. Typically, the flexibility control device seeks to maintain a mechanical compliance relationship between the environment (e.g., bulk rock) and the scoop tool so that the speed of the scoop tool changes to achieve a given reaction force. For example, in one embodiment, the flexibility control device can convert a force signal into a change in bucket movement (for example, a predetermined speed, as shown in FIG. 3C). The sensor signal input to the scoop control device can be one or more parameters selected from the cutting height, angle, predetermined force acting on the boom, characteristics of the throttle controller, preset forces acting on the boom and bucket during scooping, and gain of control devices boom and bucket for compliance, separation state and height and angle of weighing, or may include all of the parameters, including the height of the plunge, the angle, the specified force acting on the boom, the characteristic of the throttle of the regulator, given the forces acting on the boom and the bucket when digging, the gains of the boom and bucket control devices for compliance, the state of separation and the height and angle of the weighing. For example, in one embodiment, with an increase in the force exerted on the bucket, the speed of the bucket is adjusted so that the measured forces are within the specified values. The use of flexibility control devices provides embodiments that are relatively invariant to interactions between the bucket and the bulk of the rock, since they control the force exerted on the bucket and not the position of the bucket. This dynamic control of force is especially desirable for scooping through a pile of rock with random sizes of rock fragments, since it will be difficult to follow predetermined (i.e. static) predetermined paths provided that there are arbitrary obstacles that may exist in a typical rock bulk. The parameters of the compliance control device may include components of the laws of proportional, integral or differential control, and the controller may implement a linear or non-linear control circuit, for example, in accordance with a mathematical operation or function and / or in accordance with a constant. The flexibility control device may be controlled by using scoop control logic 22, such as the logic shown in the embodiment of FIG. 3B.

The activity of the compliance control device may be determined by one or more parameters. In one embodiment, these parameters are the position of the insertion throttle controller provided in the autonomous truck and the setpoints of 12 forces for the flexibility control devices 20A for the boom and bucket. However, the effectiveness of excavation is determined by the parameters of the regulator and the unknown parameters of the bulk of the rock (in a more general sense, environmental parameters). Unknown rock bulk parameters may include, but are not limited to, the size distribution of rock fragments, the shape of the bulk, rock parameters (shape, Young's modulus, Poisson's ratio, etc.), moisture content, cohesion, and angle of repose. In practice, it would be impractical to measure each of these parameters because of their quantity and because the bulk of the rock changes so often. Although equations exist for modeling more homogeneous materials, such as soils, they are generally poorly suited for modeling heterogeneous and highly changing targets, such as bulk rock of crushed rock.

The flexibility control device overcomes this problem by considering the bulk of the rock as an unmodeled object that provides a change in the reaction forces when the bucket passes through the bulk. The compliance control device uses these forces to modify the movement of the bucket without clear information about the rock bulk characteristics. While malleability control devices work well when the parameters of the malleability controllers have been tuned for the current state of the bulk pile, they may need to be reset when the bulk pile changes significantly. For example, a malleability control device configured for bulk wet rock may be too “pushy” when the bulk dries, resulting in wasted effort and reduced efficiency.

In some applications or situations, the boom actuator may be used to measure scooping force and provide a sensor signal that is used by a control / regulation device (e.g., a malleability control device, an adaptive malleability control device) to change the bucket speed. In this case, no commands are issued to the boom actuator, and the forces measured in the boom actuator depend on the interaction between the scooping tool and the bulk of the rock. Thus, the reaction forces perceived by the boom actuator characterize the interaction between the scooping tool and the bulk of the rock. In this example, the bucket actuator receives commands from the control device, and the boom actuator becomes the drive element, since no commands are transmitted to it by the control device. Alternatively, or in combination with the foregoing, a sensor signal can be obtained by measuring strain in a drive element, such as an arrow.

It should be understood that one or more other elements of the autonomous loading vehicle (i.e., elements other than the boom) can be used together with the boom actuator or, instead of the boom actuator, to provide the signal (s) of the sensor (s) to the control device and are used to control an autonomous truck, provided that such one or more other elements interact with the corresponding (appropriate) sensor (s) to generate a signal (s taps) of the sensor (s) related to the scooping parameter, such as the interaction between the scooping tool and the bulk of the rock.

In certain applications, it may be desirable to optimize the control of an autonomous truck to fill the bucket to a level less than the maximum. Such an embodiment may, for example, provide a reduction in the deformation of the components of the autonomous truck, resulting in reduced downtime for maintenance and repair and associated costs. This can be achieved by adjusting the payload based on the separation state parameter or by modifying such a parameter.

When applied in the development of deposits, the parameters of the bulk of the rock can, for example, vary significantly from one scooping to the next, even if the extracted material remains the same type of material. It has been proven that compliance regulation, such as described herein, is easily adaptable to such changes; however, significant changes in scooping conditions may result in the need to retune the compliance control device. Permanent adjustment of the compliance control device would be impractical or undesirable. Some embodiments, such as those described herein, avoid tuning problems by turning on at least one iterative learning controller (ILC) 70, 72, as shown in FIG. 3D and 3E. The iterative learning controller modifies the signals received at the input of the control device for compliance, so that the parameters of the control device can remain constant while changing the output signal of the control device. For example, as shown in FIG. 3D and 3E, an iterative learning controller can modify the sensor signal, such as the force supplied to the input of the compliance control device, so that a predetermined digging / notching mode is achieved, while simultaneously the position of the throttle infeed regulator and the predetermined forces remain the same. In one embodiment, for example, of FIG. 3E, the extent to which forces are modified is determined by the history of force deviations from previous scooping attempts.

For example, if the scoop attempt is more difficult than the previous attempt, the deviation of the force can be large and negative. In this case, a large positive corrective force will be added to all the forces introduced into the compliance control device, and the compliance control device will react more actively. The advantage of an iterative learning regulator is that it allows the algorithm to respond to a change in the state of bulk rock without the need for reconfiguration (for example, choosing constants that optimize performance) of the compliance control device. This feature saves time and eliminates the need for a specialist who would otherwise be needed for the migration process.

In another embodiment, an adaptive compliance control device may be used. Parameters (for example, components of the law of proportional, integral or differential regulation) can be adjusted or adapted dynamically (for example, in real time or essentially in real time) to compensate for rapid changes in the bulk characteristics of the rock, such as stiffness, during scooping, resulting in Avoid the need to model bulk rock. For example, an adaptive compliance control device may use force tracking deviation to dynamically control compliance parameters throughout real-time scooping. In a further embodiment, an adaptive compliance control device may be used with at least one iterative learning controller.

One embodiment of the scoop control device shown in the block diagram of FIG. 3D includes two compliance control devices 20A and two iterative learning controllers 70, 72. An additional embodiment may also include control devices with predetermined plunge and exit scenarios. A further embodiment may include a detector to determine if / when the self-contained truck is stuck. The plunge control device moves the boom and bucket actuators to the plunge position (for example, the bucket will be flush with the ground and slightly above the ground) by using, for example, a proportional position controller. After that, the autonomous loading vehicle is instructed to move towards the bulk of the rock at a speed determined by the setting of the throttle infeed regulator, and the bucket enters into contact with the bulk of the rock. When the measured forces exceed the set value, the compliance control devices will begin to move the boom and bucket actuators. When the bucket reaches its maximum tilt, the output control device takes control. The output control device moves the boom and bucket to the weighing position (for example, the bucket is fully tilted and raised above the bulk of the rock) using, for example, a proportional position controller. When the material in the bucket is weighed, the weight, scooping time and work performed by the actuators are used to measure the success of the scooping attempt. Optimum scooping maximizes bucket payload while minimizing scooping time and work. In one configuration, malleability control devices and position controllers operate at high frequencies to perform scooping operations, while iterative learning controllers operate only once per scooping cycle.

The control device for compliance can implement any mathematical dependence that provides the conversion of the range of deviations of forces into the range of possible speeds of actuators. The control device for compliance may modify the parameter, for example, depending on the magnitude of the received signal minus a given magnitude of the signal. FIG. 3E is a block diagram of a compliance control device 20A in accordance with one embodiment. The perturbed forces are used by the flexibility control device to transmit changes in the speeds of the boom and bucket actuators. The speed data is integrated to obtain a plurality of preset positions for the position controllers 60 of the boom and bucket actuators. Preset positions are monitored by position adjusters to provide the desired change in the length of the actuator. Changing the length causes the bucket to move into the bulk of the rock, which causes a change in the reaction forces. This change in force is measured by pressure sensors 90 and is used to calculate new forces 80 acting on the actuators of the boom and bucket. In one embodiment, these adjusted forces are again perturbed by an iterative learning controller designed to control compliance, before being transferred back to the compliance control device. As soon as the scooping attempt is completed, a total deviation of 85 forces is used to correct the force disturbance for the next scooping attempt, while the iterative learning controller designed to control the throttle plunge controller adds a new plunge deviation to the previous plunge deviations, so that the following disturbance / change can be calculated for the throttle infeed regulator.

In the embodiment of FIG. The 3E iterative learning controller applies a correction 87 to the default position of the throttle plunge controller based on the slope of the plunge force curve obtained from several previous scooping attempts. Other parameters used are the specified forces when scooping. As shown in the embodiments of FIG. 4A, 4B, 4C, and 4D, iterative learning controllers apply correction 92 to measured forces 94 based on a total deviation of 96 forces from several previous scooping attempts. For example, as shown in FIG. 4C, while the insertion throttle controller is initially set in accordance with the setpoint, the position of the insertion throttle controller is changed by an iterative learning controller to increase the degree of constancy of scooping efficiency. In FIG. 4C and 4D, the initial increase in force for each scooping attempt can be represented by the slope of the line passing through the smallest measured value of force and the largest measured value of force during the plunge period (between bucket plunging and compliance adjustment). These slopes are compared with the ideal slope of the plunge force curve to calculate the deviation of the slope of the curve for each scooping attempt. The deviation data is stored in memory, and their specific number n is summed. Sum 98 is used by an iterative learning controller designed to control the plunge, to calculate how the position of the plunge throttle regulator should be changed for the next scoop attempt.

As shown, for example, in FIG. 4A, an iterative learning controller can modify the input forces so that the compliance control devices respond more actively. For example, FIG. 4C and 4D show that iterative learning regulators increase a given cross section of a throttle infeed regulator and artificially increase the input forces. Increased values provide for the "introduction" of an autonomous loading vehicle into the bulk pile at a higher speed and faster tilt and lift of the bucket. In these examples, an increase in the speed of plunge, boom and bucket provides an increase in the overall activity of the scoop control device and a decrease in scooping variability compared to using parameters obtained from a pile of rock intended for training.

Embodiments of a scoop control device may be implemented on analog and / or digital (hardware / software) platforms. Certain implementations may be provided for compatibility with existing control systems, autonomous loading vehicles, sensors, etc., such as may be necessary to modernize or improve existing systems and autonomous loading vehicles. For example, a scoop control device can be implemented in whole or in part through the use of discrete elements, through the use of digital technology (for example, in the form of a digital signal processor (DSP), programmable gate arrays (FPGA) or problem-oriented / application integrated circuit (ASIC) ) or the use of their combinations. One or more components of the scoop control device may be implemented as an algorithm by using an appropriate hardware language, such as for example High Definition / High Speed Integrated Circuits (VHSL) Hardware Description Language, Inter-Register Transfer Language (RTL), or Verilog. Such an algorithm can be implemented, for example, in a device based on a programmable gate array (FPGA) or a problem-oriented / applied integrated circuit (ASIC) or other suitable logic device. Some embodiments and implementations may include one or more sensors or transducers.

Embodiments will be further described by way of the following non-limiting examples.

EXAMPLE 1

This example illustrates the design and on-site testing of one embodiment of a loading algorithm based on flexibility control using forces measured during bucket-rock interactions to modify the bucket speed during scooping. In this example, the loading algorithm (shown below) has three parts corresponding to the three phases of scooping. Three phases of scooping, namely plunging, scooping and tearing, are shown schematically in FIG. 1B, 1C and 1D.

LOAD ALGORITHM

one: procedure INJECTION (position, strength) 2: while not in the plunge position do 3: proportional regulator provides
boom and bucket movement four: end while 5: while less than the specified plunge force do 6: move forward 7: end while 8: end procedure 9: procedure SCRATCHING (forces, extensions) 10: while there is no separation state and no jam do eleven: keep moving in bulk 12: flexibility control devices
adjust the speed of actuators 13: end while fourteen: while there is a do fifteen: switch to neutral 16: expect 17: switch to forward eighteen: end while 19: end procedure twenty: procedure PERFORMANCE (position) 21: while not in do position 22: proportional regulator provides
boom and bucket movement 23: end while 24: expect steady state 25: weigh the payload in the bucket 26: end procedure

The insertion phase is shown in FIG. 1B. The insertion phase ends when the bucket is in the insertion position, and moving the autonomous loading vehicle forward causes an increase in the reaction forces of the rock acting on the bucket to values exceeding a predetermined value. During the scooping phase (Fig. 1C), the flexibility control device tilts the bucket up or down to maintain a given reaction force, while the boom is used only to measure the reaction forces when scooping. The separation phase (FIG. 1D) begins when the bucket is fully tilted and ends when the bucket is in the weighing position. A compliance control device is part of an algorithm that controls the movement of a bucket through a pile of rock. The pliability control device uses the discrepancy between the measured reaction forces when scooping and the predetermined force when scooping to change the speed of the bucket actuator. A generalized block diagram for a compliance control device is shown in FIG. 3C. While any controller C can be used to convert force deviations into actuator speeds, the flexibility control device in this example is one-way and proportional, so

(1)

(one)

where V _A is the speed of the actuator, k _A > 0 is the proportional gain (for compliance adjustment), and the deviation of the force is given by the specified force f _T minus the reaction force f _S when scooping. For this loader and bulk rock, fr was set to 80 kN, while k _A was set to 1.1 × 10 ^-7 . These values were determined experimentally by adjusting them until the mass of the payload in the bucket became large and constant. Values can also be determined using off-line tuning methods for the compliance control device, wherein the values are calculated based on known vehicle parameters and average rock hardness. The control device was limited to making only positive changes to the bucket speed (that is, only tilt up was allowed). In this example, this restriction was imposed so that no energy was expended on crushing the rock to the underside of the bucket (and maintaining the wheel drive of the autonomous truck). However, this limitation is not always necessary and may be excluded in other embodiments.

Commands for the drive mechanism were set such that the loader was driven directly into the bulk rock at a constant speed. The insertion position was chosen so that the bucket scrapped along the asphalt base to guarantee the penetration of the bucket into the rock bulk during insertion. The combination of commands for the drive mechanism and the insertion position resulted in a substantially constant penetration depth. After insertion, the throttle controller was moved to the fully open position to maximize the speed and power of the bucket actuator, while the forward thrust / forward thrust was limited by applying partial braking. The degree of braking was set such that the direct thrust caused a tendency to increase the forces acting on the actuators, which led to attempts by the flexibility control devices to reduce the forces due to the backward inclination. Direct thrust tends to “shift” the control device toward the separation state, which ensures that the scooping is completed before the forces acting on the bucket increase to values sufficient to overcome the allowable load on the actuators.

DEVICE AND METHOD OF TESTS IN PRODUCTION CONDITIONS

An automated ground loader with a lifting capacity of 1 ton and a pile of rock repelled by limestone, were used to test and configure the loading algorithm. The loader was a Kubota R250s loader, which was retrofitted for automation by adding sensors, actuators and on-board computer systems. For this example, only boom and bucket extension sensors and pressure sensors were used. The extension of the boom and bucket actuators was measured at 10 Hz using a custom Hall effect sensor. Each extension / extension sensor contained two Honeywell SPSL225 non-contact IP69 linear encoder encoders mounted in a custom housing. Two pressure sensors of the Measurement Specialties MSP-400 model were installed in the holes of the stem and cylinder of each actuator so that the resulting force acting on the actuators could be calculated. Data from pressure sensors was received at 107 Hz using one Arduino Uno board. Messages about actuator extension and pressure generated by the Arduino Uno board were transmitted to the central computer via the Robot Operating System (ROS) Electric network. The central computer was a Mini-ITX personal computer with a 64-bit Intel Core i5 processor for operation under Ubuntu 11.10 and ROS Electric. The central computer used the ROS network to publish and subscribe to information files wirelessly. The autonomous loading algorithm worked on a separate laptop with a 64-bit Intel Core i5 processor (for working under Ubuntu 11.10 and ROS Electric) connected to a wireless network. This laptop was also used to collect data.

In the tests, the full saturation control device was first used, after which the unsaturated compliance control device was used. One purpose of these tests was to determine the difference between the incline at maximum speed and the use of a compliance control device to match the incline speed with the measured reaction forces. Once the loading algorithm was set up for a given loader and bulk rock, various parameters of the loading algorithm were kept constant for scooping attempts, illustrated in the graph of FIG. 5.

Each scooping attempt began by moving the loader manually in the position in front of the bulk of the rock. The loading algorithm "took over" the control of the loader for cutting, scooping and weighing. The loader was left in the weighing position for 5 seconds to allow the bucket and payload to stop moving. The payload mass mp was calculated using equation (2) and the force f _boom acting on the boom in the weighing position. Equation (2) was determined experimentally by calibrating the loader using known masses.

(2) $$m_p=\frac{{\text{A}}_{oom}-3.9\times {\text{IP}}^{\text{four}}}{\text{51}\text{,8}}$$

(2)

Rock volume in the bucket was also checked by using video broadcast from a wireless workstation. Then the payload was dumped by manual control, and the loader was again placed in a position in front of the bulk of the rock.

EXPERIMENTAL RESULTS

A total of 57 scooping attempts were made, and the scooping efficiency for each scooping is shown in FIG. 5. Of these, 23 scooping attempts were made by tilting the bucket at maximum speed after plunging (“saturated” / maximum scooping), and 21 scooping attempts were made using a proportional (P) pliability control device, and 13 scooping attempts were made by the use of a proportional-integral (PI) flexibility control device to match the bucket speed with the measured reaction forces (adjustable scoops). Five “saturated” scoops, two scoops with proportional control and two scoops with proportional-integral control failed because the cutting forces were insufficient so that the compliance control devices with or without saturation took control of scooping. The “saturated” scooping 18 (SI8) and scooping 11 with proportional control (PI1) took 30 seconds more compared to the average value, respectively, due to problems with the wireless network and small penetration into the rock pile. Scoops, regulated by the proportional-integral control device, provided a 9% increase in payload and took 2 s more than scoops when fully saturated. Shooting attempts are shown with medium saturation (SI 1), proportional control (P7) and proportional-integral control (PI68), followed by slow scoops (SI 8 and PI 1), scooping (PI71) during heavy work and two failed scoops ( S8 and P6).

When considering the positions of actuators, valve control commands (based on valve positions) and the forces acting on the actuators, for rated scoops (SI 1 and P7), the valves on the loader have a dead zone between ± 0.5. No fluid could pass into the actuators at any position of the controlled valves within the dead zone, therefore, any command within the dead zone can be considered as a zero valve offset (closed valve). “Rich” scooping causes the bucket to tilt at maximum speed, and forces fluctuate greatly. With adjustable scooping, the flexibility control device changes the inclination speed depending on the changing forces, which leads to less abrupt changes in the forces.

For slow scoops (SI 8 and PI 1), the lack of reaction of the actuator to commands with the valve fully open and the large forces (significantly exceeding the set forces both when plunging and scooping) when trying SI8 indicated a network communication problem between the loader computer and the laptop on which the loading algorithm works. The force profile shows the level of force applied to the bulk by means of a drive mechanism when the bucket stops moving, and the final payload mass indicates that the bucket was full by the end of the scoop. The attempt of PI1 is more interesting, since the forces are very close to the given force of 80 kN during the whole scooping. When the forces increase to values exceeding the set values, the flexibility control device tilts the bucket, which causes a decrease in forces and allows the vehicle to penetrate deeper into the bulk. When the truck stops at the bulk, the forces increase and another tilt command is transmitted.

Retrospective data on commands and forces for failed scoops (S8 and P6) show that the forces never increased to values exceeding the plunging force, so that the bucket control device was never actuated and the scoops failed. Most likely, the forces were small due to the fact that the bucket hit the ground and the front wheels were unloaded. Unloading the front wheels reduces traction and reaction forces when scooping.

Work and scooping time were calculated between plunging (after reaching a predetermined plunging force) and detachment (when the bucket tilts after the detachment state). “Instant” work (only the current measured value of the force is used) was calculated using equation (3), in which Wd represents the work performed by the actuators, and Fh and F _C represent the forces when lifting and tilting, respectively, in the actuators of the boom and bucket, and dh and d _C represent offsets for each actuator. Let n be the total number of sensor readings and let the subscript i denote the time index associated with each sensor read. Thus, instant work was calculated as:

(3) $$W_d=\frac{\mathrm{one}}{2}{\displaystyle \sum _{i=\mathrm{one}}^{n-\mathrm{one}}[F_{h,i}}\cdot |\; |\; |d_{h,i}-d_{h,i+\mathrm{one}}|\; |\; |+F_{c,i\cdot }\cdot |\; |\; |d_{c,i}-d_{c,i+\mathrm{one}}|\; |\; |]$$

(3)

The average values of the scooping efficiency are shown in Table 1 for both “saturated” scoops and for scooping with proportional control. The scooping time increased sharply when the bucket was controlled by means of a proportional control device for compliance, compared to when the bucket was moved at its maximum speed. However, all regulated scoops were more consistent than “saturated” scoops, and the payload mass increased by 10%.

TABLE 1 Average values of the efficiency of “saturated” scoops and scoops with proportional control ε _d “Saturated” Adjustable Difference t _d [s]
W _d [N m]
M _d [kg] 7.0 ± 12.3
729 ± 176
654 ± 384 24.3 ± 11.2
444 ± 107
721 ± 170 + 247%
-61%
+ 10%

The integral work was calculated by using equation (4), which uses both the current force reading and the last force reading in the integral work equation:

(4) $$W_d=\frac{\mathrm{one}}{2}{\displaystyle \sum _{i=l}^{n-\mathrm{one}}\left[(F_{h,i}+F_{h,i+\mathrm{one}})\cdot |\; |\; |d_{h,i}-d_{i,i+\mathrm{one}}|\; |\; |+(F_{c,i}+F_{c,i+\mathrm{one}})\cdot |\; |\; |d_{c,i}-d_{c,i+\mathrm{one}}|\; |\; |\right]}$$

(four)

The results are presented in Table 2.

TABLE 2 Average values of the efficiency of “saturated” scoops and scoops with proportional and proportional-integral regulation All scoops ε _d “Saturated” Difference Proportional control device Difference Proportional-integral control device t _d [s]
W _d [N m]
M _d [kg] 6.5 + 12.1
24458 ± 8495
721 + 326 + 277%
+ 15%
0.2% 24.5 + 11.3
28145 + 6711
719 ± 170 -33%
-3%
+ 10% 8.1 + 5.0
27287 + 14511
792 + 335 No failed scoops t _d [s]
W _d [N m]
M _d [kg] 6.2 + 12.6
25658 ± 8697
784 + 254 + 315%
+ 15%
0.2% 25.5 ± 10.1
29551 ± 1929
755 + 40 -37%
+ 8%
+ 22% 9.0 + 3.8
32246 + 8760
924 + 85

EXAMPLE 2

Key parameters for the insertion and scooping phases performed by the autonomous material handling vehicle can be identified as the position of the insertion throttle controller and the predetermined forces acting on the boom and bucket. Target force profiles, such as those shown in FIG. 6A and 6B can be used as a basis for dividing an iterative learning regulator to control loading into two parts: an iterative learning regulator for plunge control, which controls the throttle plunging regulator, and an iterative learning regulator for a pliability control device that modifies the measured forces, introduced into the compliance control devices.

ITERATIVE TRAINING CONTROLLER FOR CUTTING CONTROL

The iterative learning controller for plunge control compared the slope of the plunge force profile acting on the boom with the slope of the set plunge force profile, as shown in FIG. 7. An increase in forces to values less than 100 kN / s indicates that the bulk of the rock created less resistance compared to the bulk of the rock intended for training, while an increase in forces to values above this specified value indicates a bulk with higher resistance. Correction for bulk with lower resistance can be achieved by adjusting the throttle infeed regulator in accordance with the relationship shown in equation 5.

_{наклон при врезании} (5)C _{Insertion throttle control} = $${\Gamma }_{\mathrm{at}Res\mathrm{but}n\mathrm{and}e}\cdot {\displaystyle \sum _{i=\mathrm{one}}^n{\tau }_{y}3/\mathrm{four}}*\mathrm{<figure-callout\; id="3"\; label="e"\; filenames="00000009.png"\; state="\{\{state\}\}">e</figure-callout>}3/{\mathrm{four}}_{}$$

_{insertion angle} (5)

In equation (5), Γ _insertion is any given control device that converts the deviation e of the _{incline during insertion} into the opening increment of the throttle controller, referred to as correction C _{The insertion} throttle controller for the _insertion throttle controller. The simplest control device is a proportional infeed control device that linearly converts the inclination of the profile into the corrective increment of opening the throttle controller. i represents the number of digging cycles to be considered, w _i is the possible weight assigned to each deviation of the profile inclination during insertion. This weight value can be used to offset the correction to a given set of indications of deviations of the slope of the profile during insertion. For example, the most recent deviations of the slope of the profile during insertion are probably the best way to characterize the current state of bulk rock. For example, a weight that decreases exponentially with respect to i can be used to assign the largest weights to the most recent tests.

Equation 6 shows a specific example of an iterative learning controller for plunge control, in which the profile deviations for five scoop attempts are multiplied by an exponentially decreasing weight and are summed before multiplication by a proportional plunge gain.

_{наклон при врезании} (6)C _{Insertion throttle control} = $${\gamma }_{\mathrm{at}Res\mathrm{but}n\mathrm{and}e}\cdot {\displaystyle \sum _{i=\mathrm{one}}^5\frac{\mathrm{one}}{i^2}}{e}_i$$

_{insertion angle} (6)

After five additional plunge attempts, the initial adjustment will be forgotten, and the current adjustment will be based on five later profile tilts during insertion. A similar iterative learning controller is used to apply corrections to the measured forces input to the compliance control devices.

ITERATIVE TRAINING REGULATOR FOR RESISTANCE

The iterative learning regulator for compliance control compares the calculated forces with the given forces and uses the result to modify the forces used by the compliance control devices. An exemplary force profile is shown in FIG. 8. The total discrepancy e _{F Net} between the measured forces and the given forces is calculated using equation 7.

(7) e _{F Net} = $${\displaystyle {\int }_{t_{n\mathrm{but}h\mathrm{but}l\mathrm{about}}}^{t_{\mathrm{to}\mathrm{about}nec}}({\mathrm{<figure-callout\; id="3"\; label="F"\; filenames="00000009.png"\; state="\{\{state\}\}">F</figure-callout>}}_{3\mathrm{but}d\mathrm{but}nn\mathrm{but}\mathrm{I\; am}}}-F_{\mathrm{AND}smeRenn\mathrm{but}\mathrm{I\; am}})\cdot dt$$

(7)

And in this case, the integral of the force difference (F _Preset - F _Measured ) is taken according to the duration of the scooping between t _beginning and t _end . Correction C _{The measured forces of the} measured forces, determined by the iterative learning controller for compliance control, are calculated using equation 8.

⋅ ϒ_{податливость} ⋅ e _{F Net} (8)C _{Measured forces} = $${\displaystyle \sum _{i=\mathrm{one}}^5\frac{\mathrm{one}}{{\mathrm{<figure-callout\; id="2"\; label="i"\; filenames="00000009.png,00000023.png"\; state="\{\{state\}\}">i</figure-callout>}}^2}}$$

⋅ ϒ _compliance ⋅ e _{F Net} (8)

And in this case, the exponentially decaying term is used to shift the adjustment to the five most recent attempts at scooping, but any number of scooping attempts can be taken into account, and all of them can have the same weight. ϒ _Compliance is the gain of an iterative learning regulator to control compliance. Any conventional controller / regulator with a gain ϒ _compliance can be used instead of a proportional controller with a gain ϒ _compliance .

The advantage of iterative learning regulators is that after the parameters of the malleability control device are configured for a given vehicle and rock pile, they will never need to be tuned again. Each of the iterative learning controllers discussed in this section has only two parameters: the number i of previous scooping attempts and the gain coefficients of iterative learning controllers. Another way of considering the amplification factors of iterative learning regulators is to consider, from the point of view of activity, the entire scooping algorithm (a compliance control device and an iterative learning regulator).

If the amplification factors of the iterative learning regulators are large, the compliance control devices will react more actively to changes in the bulk of the rock, and if the amplification factors of the iterative learning regulators are small, the regulation devices will respond less actively. This level of regulation is ideal for operators, since one value can be set based on general loading goals. If the payload of the truck is lower than the specified mass flow rate of the mill / crusher, the operator can increase the activity of the control device by increasing the gain of the iterative learning controllers. If the mass flow exceeds that which the mill / crusher can handle, the gain of the iterative learning controllers can be reduced, for example, to save on tire wear and fuel consumption.

The contents of all references cited herein are hereby expressly incorporated herein by reference.

EQUIVALENTS

Those skilled in the art will be able to identify or identify variations of the embodiments described herein. Such varieties are within the scope of the invention and are covered by the appended claims.

Claims (38)

1. A scoop control device for an autonomous material handling vehicle (ALV) used in the mining and construction industries, comprising: at least one controller controlling the bucket and / or autonomous material handling vehicle in accordance with a loading algorithm based on compliance control using at least one sensor signal, wherein at least one sensor signal characterizes the interaction between the bucket and the bulk of the rock during scooping; wherein at least one sensor signal is obtained from at least one sensor interacting with one or more actuators other than the bucket actuator, or with one or more drive elements, wherein at least one sensor signal is obtained by measuring force, perceived by the actuator boom, and at least one controller is configured to convert a signal characterizing the force into a change in the movement of the bucket. 2. The scoop control device according to claim 1, wherein the at least one controller comprises at least one flexibility control device. 3. The scoop control device according to claim 1, wherein the final payload is adjusted by modifying a parameter corresponding to the separation phase during scooping. 4. The scoop control device according to claim 1, wherein at least one sensor signal is obtained by measuring the force perceived by the drive element. 5. The scoop control device according to claim 2, wherein the compliance control device comprises an adaptive compliance control device. 6. The scoop control device according to claim 5, wherein the adaptive flexibility control device dynamically adjusts at least one parameter in response to a difference between the sensor signal and the predetermined signal. 7. The scoop control device according to claim 1, wherein the at least one controller converts one or more sensor signals into a range of possible bucket speeds or speeds of the autonomous truck by using at least one of proportional, integral, and differential control. 8. The scoop control device according to claim 1, comprising at least one iterative learning controller (ILC), which uses feedback from at least one previous scooping to modify at least one sensor signal provided to at least one controller. 9. A scoop control device for an autonomous material handling vehicle (ALV) used in the mining and construction industries, comprising: at least one controller controlling the bucket and / or autonomous material handling vehicle in accordance with a loading algorithm based on compliance control using at least one sensor signal, wherein at least one sensor signal characterizes the interaction between the bucket and the bulk of the rock during scooping, at least one sensor signal is obtained by measuring the force perceived by the boom actuator; and at least one iterative learning controller (ILC), which uses feedback from at least one previous scooping to modify at least one sensor signal provided to at least one controller, wherein at least one controller is configured to convert a signal indicative of force into a change in bucket movement. 10. The scoop control device according to claim 9, wherein the at least one controller comprises at least one flexibility control device. 11. The scoop control device of claim 10, wherein the compliance control device comprises an adaptive compliance control device. 12. The scoop control device according to claim 9, comprising a first iterative learning speed controller that changes the speed of the autonomous loading and sufficient vehicle based on a sensor signal characterizing the interaction between the bucket and bulk of the rock during at least one previous scooping, and a second iterative learner a regulator that modifies the sensor signal obtained from measuring the deviation of the force acting on the bucket and boom with at least one previous scooping. 13. The scoop control device according to claim 9, including a first iterative learning controller modifying the sensor signal outputted by the malleability control device, and a second iterative learning controller modifying the sensor signal outputting the malleability adjustment device, the modification being based feedback from at least one previous scooping. 14. The scoop control device according to claim 9, wherein the at least one iterative learning controller converts the signal from the previous scooping into changes in the output of the scoop control device by using at least one of proportional, integral, and differential control. 15. A non-volatile storage medium compatible with a computer for use with a scoop control device for a self-contained material handling vehicle and comprising a computer, the medium storing a computer program containing computer control instructions for executing one or more of: receiving at least one sensor signal from at least one sensor interacting with one or more actuators other than the bucket actuator or one or more drive elements; wherein at least one sensor signal characterizes the interaction between the bucket and the bulk of the rock during digging, and at least one sensor signal is obtained by measuring the force perceived by the boom actuator; converting the signal characterizing the force into a change in the movement of the bucket and controlling a bucket and / or autonomous truck in accordance with a loading algorithm based on compliance control using at least one sensor signal. 16. A method of controlling an autonomous material handling vehicle (ALV) used in the mining and construction industries, in which they carry out: receiving at least one sensor signal from at least one sensor interacting with one or more actuators other than the bucket actuator or one or more drive elements; wherein at least one sensor signal characterizes the interaction between the bucket and the bulk of the rock during digging, and at least one sensor signal is obtained by measuring the force perceived by the boom actuator; converting the signal characterizing the force into a change in the movement of the bucket and control of a bucket and / or autonomous truck in accordance with a loading algorithm based on flexibility control using at least one sensor signal. 17. The method according to p. 16, in which at least one sensor signal is obtained by measuring the force applied by the boom actuator, or by measuring strain in the drive element. 18. The method according to p. 16, in which at least one sensor signal output to the control device / controller is modified so that the speed of the bucket or the speed of the autonomous truck changes. 19. The method of claim 16, wherein the control further includes dynamically adjusting at least one parameter in response to a difference between the sensor signal and the predetermined signal. 20. The method according to p. 16, in which additionally control at least one of the movement of the bucket of the autonomous truck to at least one selected position and the movement of the autonomous truck relative to the bulk of the rock. 21. The method according to p. 16, in which additionally provide a change in the speed of the autonomous truck based on the sensor signal characterizing the interaction between the bucket and the bulk of the rock during at least one previous scooping, and modify the sensor signal obtained from the measurement of deviation the force acting on the bucket and boom with at least one previous scooping. 22. The method of claim 16, further comprising modifying the sensor signal outputted by the malleability control device and modifying the sensor signal outputted by the malleability adjustment device, the modification being based on feedback from at least one previous scooping. 23. The method according to p. 16, in which the control includes converting the total deviation of the force into a range of possible measured forces by using at least one of proportional, integral and differential regulation. 24. The method of claim 16, further comprising converting the signal from previous scooping to changes in the output of the scoop control device by using at least one of proportional, integral, and differential control. 25. The method according to p. 16, in which the final adjustment of the final payload is carried out by modifying the parameter corresponding to the separation phase during scooping.

Images