Method for determining the setpoint values ​​of a manipulated variable for a pressure medium supply system of a hydraulic system

Abstract

The invention relates to a method for determining control values ​​of a manipulated variable for a pressure medium supply system of a hydraulic system with at least one hydraulic consumer (10), wherein the pressure medium supply system comprises a hydraulic machine (2) and an electric drive (4, 6) coupled thereto and is controlled by the manipulated variable, wherein an inverse plant model (47) is given for the hydraulic system, which uses values ​​of a state variable to be controlled of the hydraulic system and of time derivatives of the state variable up to a predetermined maximum order as input values ​​and maps these to an output value for the manipulated variable; comprising determining (110) trajectory values ​​(36) of a trajectory for the time course of the state variable and derivative values ​​(38, 40) for time derivatives of the course of the state variable up to the maximum order from predetermined setpoint values ​​(32) for the state variable;Determining (120) compensation values ​​(73) with a compensation controller (75) from a control deviation (61) between the trajectory values ​​(36) and actual values ​​(54) of the state variable; and determining (130) manipulated values ​​(50) for the manipulated variable using the inverse plant model (47) from the trajectory values ​​(36), the derivative values ​​(38, 40) and the compensation values ​​(73).

Description

[0001] The present invention relates to a method for determining control values ​​of a control variable for a pressure medium supply system of a hydraulic system, as well as a computing unit and a computer program for its implementation and a hydraulic system. Background of the invention

[0002] Machines, especially mobile construction equipment (e.g., excavators), can have a working hydraulic system, i.e., a hydraulic system, to effect the movement of elements (e.g., boom sections) by means of hydraulic actuators (e.g., hydraulic cylinders and / or hydraulic motors). One way to make the use of the working hydraulic system more efficient is to provide a demand-based flow rate, for example, by using a variable displacement pump, i.e., a hydraulic pump with a variable displacement, which, however, are typically relatively expensive compared to fixed displacement pumps. In hydraulic systems that have a fixed displacement pump driven by an electric motor, the speed of the fixed displacement pump or the electric motor can be varied to vary the delivered flow rate.To supply hydraulic systems used in mobile applications to control work equipment with the desired flow rate, various control concepts are possible. For example, in flow-rate-based systems such as EFM (Electro-hydraulic Flow Matching), flow rate or speed control is used. In systems whose control is based on pressure or pressure differential, such as LS and LUDV systems (LS: Load-Sensing, LUDV: Load-Pressure-Independent Flow Distribution), pressure control is employed. Disclosure of the invention

[0003] According to the invention, a method for determining the setpoints of a control variable for a pressure medium supply system of a hydraulic system (or electro-hydraulic system), as well as a computing unit and a computer program for its execution, and a hydraulic system with the features of the independent claims are proposed. Advantageous embodiments are the subject of the dependent claims and the following description.

[0004] In the method for determining the control values ​​of a manipulated variable for a pressure medium supply system of a hydraulic system, an inverse plant model for the hydraulic system is used, which takes values ​​of a state variable to be controlled of the hydraulic system and of time derivatives of the state variable up to a predetermined maximum order as input values ​​and maps these to an output value for the manipulated variable.The procedure includes determining trajectory values ​​of a trajectory for the time course of the state variable and derivative values ​​for time derivatives of the course of the state variable up to the maximum order from given setpoint values ​​for the state variable, determining compensation values ​​with a compensation controller from a control deviation between the trajectory values ​​and actual values ​​of the state variable, and determining manipulated values ​​for the manipulated variable using the inverse plant model from the trajectory values, the derivative values, and the compensation values.By determining the trajectory for the state variable and using the inverse plant model, feedforward control with low deviations is enabled, in particular reducing undesirable discontinuities in the controlled state variable, which can occur, for example, when the manipulated values ​​are determined directly from the specified setpoints (e.g., with a single controller).

[0005] The manipulated variable is, for example, the torque to be applied by the electric drive or the rotational speed of the electric drive. The state variable is, for example, a pressure in the hydraulic system, in particular a pressure at a connection of the hydraulic machine or a hydraulic line or channel connected to the connection, a rotational speed of the hydraulic machine, or a volumetric flow rate of hydraulic fluid delivered by the hydraulic machine.

[0006] In one embodiment, when determining the manipulated variable's values, the derivatives of the time derivative with the maximum order are modified by compensation values ​​to obtain the altered derivatives. The trajectory values, the derivatives of time derivatives with an order smaller than the maximum order, and the modified derivatives are used as inputs to the inverse system model. Specifically, the compensation values ​​are added to the derivatives of the time derivative with the maximum order to determine the altered derivatives. Alternatively, the compensation values ​​can be used to modify the output values ​​of the inverse system model (e.g., added to them), in which case the trajectory values ​​and the derivatives (of all orders) remain unchanged, or at least are not modified by the compensation values, as inputs to the inverse system model.

[0007] According to one embodiment, a filter whose order equals the maximum order is used to determine the trajectory values ​​and derivative values ​​from the specified setpoints for the state variable. The filter, in particular, comprises a number of series-connected first-order (PT1) filters or is implemented as such, with the number of filters equal to the maximum order. Using a filter is advantageous because it avoids the direct calculation (e.g., by subtraction) of derivatives from measured values ​​that are, for example, noisy.

[0008] According to one embodiment, one or more characteristic maps for the hydraulic machine are used in the inverse system model. In particular, the following are used: a volumetric flow rate characteristic map, which specifies the volumetric flow rate delivered by the hydraulic machine and which depends on a pressure and a rotational speed of the hydraulic machine; and / or a torque characteristic map, which specifies the torque occurring at the hydraulic machine and which depends on the pressure and rotational speed of the hydraulic machine; and / or one or more characteristic maps for the derivatives of the volumetric flow rate characteristic map and / or the torque characteristic map with respect to pressure or rotational speed; and / or one or more inverted characteristic maps, each of which is the inverse of the volumetric flow rate characteristic map or the torque characteristic map resolved with respect to pressure or rotational speed.Characteristic maps, in particular, allow for the consideration of leakage and / or friction, the effects of which generally cannot be expressed by simple equations. The characteristic maps can be stored, for example, in a computing unit that implements the method. The characteristic maps themselves can be stored, for example, in tabular form, or a machine learning-based model, such as a neural network trained on the characteristic map(s), can be used and stored in the computing unit. The latter can, for example, require less storage space.

[0009] According to one embodiment, when determining the trajectory values ​​and the derivative values, the derivative values ​​for one of the time derivatives, especially the one with the highest order, are limited upwards and / or downwards by respective first limit values. This prevents overdriving, e.g., due to an integral component in the compensation controller, when the specified setpoints change rapidly, thus preventing oscillations in the manipulated values.

[0010] According to one embodiment, the initial limit values ​​are varied depending on the trajectory values ​​and / or the derivative values ​​that are not limited, and / or depending on the actual value of the manipulated variable, and / or depending on a possible range of values ​​of the manipulated variable, and / or depending on estimated values ​​of at least one disturbance variable determined by a disturbance estimation method. This disturbance variable is an unobserved or unobservable quantity in the hydraulic system. The at least one disturbance variable includes, in particular, the total volume flow rate of the at least one consumer and / or its time derivative. This allows for the consideration of actual limitations, such as the available drive torque (e.g., when the drive torque is reduced to prevent overheating of the electric drive) or different operating conditions.

[0011] According to one embodiment, a disturbance estimation method is used to determine estimated values ​​for at least one disturbance variable, which is an unobserved or unobservable quantity in the hydraulic system. These estimated values ​​are then used to determine the trajectory values ​​and derivative values, and / or in the inverse system model. The at least one disturbance variable includes, in particular, the total volumetric flow rate of the at least one consumer and / or its time derivative. By considering the disturbance variable, the accuracy of, for example, the inverse system model can be increased. The disturbance estimation method can be, for example, an observer, in particular a Luenberger observer, or an iterative estimation method, in particular a Kalman filter.

[0012] According to one embodiment, the disturbance estimation method uses or evaluates actual values ​​of the torque and / or speed applied by the electric drive. These actual values ​​are typically available in the inverter control unit and can be transmitted by it or, if the method according to the invention is implemented in the inverter control unit itself, used within it.

[0013] According to one embodiment, when determining the control values ​​for the manipulated variable, the output values ​​of the inverse plant model are restricted to determine the control values. This allows, for example, the exclusion of values ​​that are technically impossible or that are impossible due to the situation. The restriction, i.e., its lower and / or upper limit, can be changed, for example, depending on the torque that can be applied by the electric drive, which may depend on the temperature of the electric drive.

[0014] In one embodiment, the compensation controller has an integral component, where difference values ​​between the output values ​​and the limited control values ​​are determined and used in a first anti-windup function of the integral component. The anti-windup function can be implemented, for example, such that the control deviation is corrected or modified by the difference value; that is, the difference value is subtracted from the control deviation (where the signs are, of course, appropriately defined so that the absolute value of the control deviation is reduced), and the corrected / modified control deviation, after possible multiplication by an integral factor, is used in an integrator of the integral component, thereby ultimately preventing the integral from growing unchecked.

[0015] In one embodiment, the compensation controller has an integral component, wherein a second anti-windup function of the integral component depends on the derivative values, in particular the derivative values ​​of the first time derivative. Specifically, the second anti-windup function limits the control deviation for the integral component downwards and / or upwards by means of respective second limit values, where the second limit values ​​depend on the derivative values, in particular the derivative values ​​of the first time derivative, and / or the second anti-windup function modifies an integral factor depending on the derivative values, in particular the derivative values ​​of the first time derivative. The second limit values ​​and / or the integral factor can be smaller in absolute value for derivative values, in particular the derivative values ​​of the first time derivative, that are larger in absolute value.This design is expedient because it prevents an undesirably high increase in the integral component at high rates of change of the specified target values.

[0016] A computing unit according to the invention, e.g., a control unit of a hydraulic system of a mobile working machine or an inverter control of an electric drive by which a hydraulic machine of a hydraulic system is driven, is configured, particularly by means of programming, to carry out a method according to the invention. A hydraulic system with a computing unit according to the invention is also the subject of the invention.

[0017] Implementing a method according to the invention in the form of a computer program or computer program product with program code for carrying out all method steps is also advantageous, as this incurs particularly low costs, especially if an executing control unit is already available for other tasks. Suitable data carriers for providing the computer program are, in particular, magnetic, optical, and electrical storage media, such as hard drives, flash memory, EEPROMs, DVDs, etc. Downloading a program via computer networks (Internet, intranet, etc.) is also possible.

[0018] Further advantages and embodiments of the invention will become apparent from the description and the accompanying drawing.

[0019] It is understood that the features mentioned above and those to be explained below can be used not only in the combinations specified, but also in other combinations or on their own, without leaving the scope of the present invention.

[0020] The invention is schematically illustrated in the drawing using exemplary embodiments and is described in detail below with reference to the drawing. Character description Fig. Figure 1 shows a hydraulic system with a constant-displacement pump driven by an electric machine, which can be operated using the method according to the invention. Fig. Figure 2 shows a flowchart according to an exemplary embodiment of the invention. Fig. Figure 3 shows a control structure according to an exemplary embodiment of the invention. Fig. Figure 4 illustrates a filter structure for determining trajectory values ​​and associated derivative values ​​from given setpoint values ​​for the state variable. Fig. Figure 5 illustrates the use of a disturbance observer or observer. Fig. Figure 6 illustrates a further embodiment which reduces overshoot, in particular, during transient transitions with abrupt behavior or with large setpoint change rates. Detailed description of the drawing

[0021] Fig. Figure 1 shows a hydraulic system or electro-hydraulic system with a hydraulic machine 2 driven by an electric drive, which is, for example, a constant displacement pump (i.e., a hydraulic pump with constant displacement V). gor constant swallowing volume, where displacement denotes the volume conveyed per revolution), which can be operated with the method according to the invention. The hydraulic machine and the electric drive together form a pressure supply system or hydraulic fluid supply system of the hydraulic system, through which hydraulic pressure medium (hydraulic fluid, in particular hydraulic oil) is provided or conveyed in the system.

[0022] The hydraulic machine 2 (hereinafter also referred to simply as the pump) has two hydraulic connections. One of the connections, referred to as the tank-side connection, is connected to a tank for hydraulic pressure medium (hydraulic fluid, in particular hydraulic oil), and the other connection, referred to as the outlet-side connection, is connected via a hydraulic line 12 to a valve arrangement 8 (e.g., a modular valve block), which is not shown in detail. This valve arrangement controls the flow of pressure medium to and from at least one hydraulic consumer 10 (here, for example, a hydraulic cylinder). The valve arrangement 8, i.e., the valves or valve spools it comprises, is controlled by a control unit 20, e.g., based on user inputs that are detected by an operating device (e.g., a joystick; not shown) and transmitted to the control unit 20.Optionally, a volume flow rate and / or pressure rate (i.e., a volume flow rate setpoint and / or a pump pressure setpoint) can also be determined by the control unit 20 based on user input.

[0023] A pressure sensor (i.e., a sensor that measures the pressure of the hydraulic fluid), designated as a load pressure sensor 14, can be provided on the hydraulic line between the valve assembly 8 and the hydraulic consumer 10. This sensor allows the load pressure of the hydraulic consumer, or the maximum load pressure in the case of multiple hydraulic consumers, to be detected and measured. In a system with direct flow rate control, a load pressure sensor can be omitted, in which case, for example, the control unit 20 can determine the flow rate and / or (maximum) pressure.

[0024] A pressure sensor or pump pressure sensor 16 is provided on the output side of the hydraulic machine 2, i.e., at the output port of the hydraulic machine or on the hydraulic line 12 between the hydraulic machine 2 and the valve block 8. The pump pressure sensor 16 measures the pump pressure, which here is considered to be the output pressure of the hydraulic fluid present at the output port of the constant-pressure pump or on the hydraulic line between the constant-pressure pump and the valve assembly. Alternatively, the pump pressure sensor can be provided directly on the hydraulic machine, for example, if it has a dedicated port where the pump pressure is supplied as a hydraulic signal.

[0025] The electric drive comprises an electric machine 4 and an inverter 6 (power converter, in particular an inverter), which supplies the electric machine 4 and its phase windings with alternating current voltages. The inverter 6 is connected, for example, via a DC link to an electrical energy source, such as a battery. The electric machine 4 is mechanically coupled to the hydraulic machine 2, or rather, rotationally fixedly coupled, for example, by means of a shaft, and a gearbox with a specific gear ratio may also be provided. The rotational speed of the hydraulic machine 2 is therefore equal to the rotational speed of the electric machine 4, or, taking the gear ratio into account, corresponds to the rotational speed of the electric machine 4; that is, the ratio of the two rotational speeds is equal to 1 or equal to the gear ratio.

[0026] An inverter control unit 22 is provided, which is configured to control the inverter 6 according to a manipulated variable, wherein the manipulated variable is, in particular, the torque to be applied by the electric machine 4 (which is caused by alternating currents generated in the inverter). Values ​​for the manipulated variable, referred to as manipulated values, can be transmitted to the inverter control unit 22 by a higher-level control unit, such as the control unit 20, which determines the manipulated values, or can be determined, in whole or in part, by the inverter control unit 22 itself. A method according to the invention, which is implemented in the inverter control unit 22 and / or the control unit 20, can be carried out to determine the manipulated values ​​to be used.

[0027] Fig. Figure 2 shows a flowchart according to an exemplary embodiment of the invention. It is a method for determining the setpoints of a manipulated variable for a pressure medium supply system of a hydraulic system (as is the case, for example, in Fig. Figure 1 shows the hydraulic system. The hydraulic system has at least one hydraulic consumer, which is supplied with hydraulic fluid by the hydraulic fluid supply system. The hydraulic fluid supply system includes a hydraulic machine and an electric drive coupled to it, and is controlled by the manipulated variable.

[0028] It is assumed that an inverse system model is given for the hydraulic system, which uses values ​​of a state variable to be controlled (e.g. pressure or volume flow) of the hydraulic system and of time derivatives of the state variable up to a predetermined maximum order as input values ​​and maps these to an output value for the manipulated variable.

[0029] The steps of the process are repeated and performed continuously, so the use of the plural for "values" (e.g., setpoints, trajectory values, derivative values, manipulated values, compensation values, and other values) refers specifically to the fact that temporal sequences of such values ​​exist and are processed. There are, therefore, values ​​for different points in time. Each value (for each of the sequences) refers accordingly to a point in time. The term "values" can thus be understood as "a temporal sequence of values, each of which refers to a point in time." When the process is implemented by a computing unit (e.g., a control unit), the set of points in time is typically discrete, e.g., a sequence of equally spaced points in time.If different values ​​(from different sequences) are processed together in the procedure, then values ​​(from the different sequences) that refer to the same point in time are processed together. The term "derivative values" is intended to include several temporal sequences, namely, where provided, one temporal sequence of derivative values ​​for each order of temporal derivatives; that is, temporal sequences of derivative values ​​are provided for the first-order temporal derivative, ... for the maximum-order temporal derivative. Alternatively, the terms "first-order derivative values," ... "maximum-order derivative values" could be used.

[0030] In an optional step 100, setpoints for the (controlled) state variable are determined. This can be done, for example, by evaluating a user signal acquired via a user interface (e.g., a joystick), whereby different levels of the user signal (e.g., corresponding to different joystick deflections) are mapped to corresponding setpoints for the state variable (e.g., including filtering and / or progressive mapping). Alternatively, the setpoints for the state variable could be specified by a higher-level controller. For the following steps, it is assumed that setpoints for the state variable are predefined, i.e., a sequence of setpoints, each referring to a specific point in time.

[0031] In step 110, trajectory values ​​for the temporal evolution of the state variable and derivative values ​​for time derivatives of the evolution of the state variable are determined from the specified setpoint values, e.g., using a state variable filter. The time derivatives are determined up to the maximum order that occurs in the inverse plant model.

[0032] In step 120, compensation values ​​are determined from a control deviation, i.e., from control deviation values, between the trajectory values ​​and the actual values ​​of the state variable using a compensation controller. The control deviation is specifically the difference between the trajectory values ​​and the actual values ​​of the state variable (i.e., trajectory value minus actual value, each referenced to the same time). The actual values ​​of the state variable are measured by a sensor (e.g., a pump pressure sensor) and transmitted (e.g., to a processing unit that implements the procedure).

[0033] In step 130, control values ​​for the manipulated variable are determined using the inverse path model from the trajectory values, the derivative values, and the compensation values. The trajectory values ​​and the derivative values ​​are used, possibly in a modified form based on the compensation values ​​(for example, the derivative values ​​of the maximum order can be changed by the compensation values, as in the Fig. 3, Fig. 5 and Fig. 6), are used as input values ​​for the inverse plant model. The output values ​​of the inverse plant model can be used directly as control values ​​or can be used in a modified or changed form as control values ​​(for example, a restriction can be provided, as shown in Fig. 3 shown).

[0034] In an optional step 140, the electric drive is controlled with the control values, or the control values ​​are transmitted to or used in a higher-level control system, which uses the control values ​​to determine the final control values ​​with which the electric drive is controlled. The higher-level control system can, for example, implement a further modification of the control values, or it can be a secondary control system, in which, according to the invention, a controller is selected from several parallel controllers, from whose respective determined (possible) control values ​​one is selected as the final control value. In a secondary control system, the control values ​​determined in step 130 could, for example, be considered possible control values ​​that can be temporarily selected from several possible control values ​​determined by different controllers.

[0035] The following Fig. Figures 3 to 6 illustrate further details and design options, especially for steps 110 to 130.

[0036] Fig. Figure 3 shows a control structure according to an exemplary embodiment of the invention. The control structure is described using the example of a pressure regulator, but is in principle also applicable to a controller that regulates a different state variable, e.g., a volume flow rate or a rotational speed.

[0037] It is assumed that setpoint values ​​32 for the state variable to be controlled are specified, e.g. by a higher-level control system. These specified setpoint values ​​32 generally vary over time.

[0038] The core of this is the inverted plant model 46 of the system under consideration, which serves as feedforward control. The inverse plant model 8 has an input for values ​​of the state variable to be controlled, where a temporal sequence of values ​​of the state variable, i.e., a setpoint trajectory or trajectory of the state variable, is given, so that the values ​​of the state variable to be controlled at the input of the inverse plant model are also referred to as trajectory values ​​36. Further inputs are for time derivatives 38 and 40 or 44 of the state variable to be controlled, i.e., the trajectory, where the number of derivatives results from the order of the system of differential equations by which the system is described. The order of the highest derivative is referred to as the maximum order. The inverse plant model 46 generates (from the values ​​at the inputs) values ​​47 for the manipulated variable (referring to Fig. 1 for example, for the torque to be applied by the electric drive).

[0039] In the case of pressure control, the state variable is the pump pressure. A state variable filter 34 serves, firstly, to filter the specified setpoint values ​​32 of the state variable, and secondly, to determine or calculate time derivatives of the trajectory, or values ​​for the time derivatives, referred to as derivative values. In the example shown, these are derivative values ​​38 of a first time derivative and derivative values ​​40 of a second time derivative.

[0040] Furthermore, it is possible to adapt setpoints 32 to a possible manifold constraint, for example, so that the feedforward control component is better utilized, especially in transient processes, and a lower control deviation is achieved, which in the case of a PI controller (controller with proportional component 64 and integral component 66, 70) could otherwise lead to increased overshoot of the controlled state variable. Since, as a rule, not all real physical effects can be correctly represented in a model (at least not with finite effort), corresponding model errors contained in the inverse plant model 46 can lead to a deviation between the setpoint, i.e., trajectory, and the actual value. A controller designated as a compensation controller 75 is provided for compensation, which is Fig. 3 is implemented as an example PI controller.

[0041] In this process, the control deviation 61 between the trajectory value 36 and the actual pressure value 54 is calculated at the difference element 60 (which refer, for example, to the same time or are available at the evaluation time) and is calculated via the Pl controller (a P controller, i.e., a controller with a proportional component, or a PID controller, i.e., a controller with a proportional, integral, and differential component, are of course also conceivable), consisting of a P factor 64 (proportional factor) and an I factor 66 (integral factor), as well as the actual integrator element 70, whereby the two components are added to a summation element 72, a compensation value 73 is calculated and added to the second derivative of the trajectory, i.e., to the derivative values ​​40, via the summation element 42 and passed to the inverse plant model as an input variable 44. The compensation component of the second derivative controller (or...)Adding the highest derivative (in the general case) is expedient, since the highest derivative of the system state always has a direct relationship to the manipulated variable. Alternatively, adding the compensation value to the manipulated variable value 47 determined by the inverse plant model 46 is also possible.

[0042] A control variable limit can be implemented using the limiting element 48, which determines the control value 50 for the control variable, for example, as a value limited both upwards and downwards from the value 47 for the control variable determined by the inverse plant model 46. Here, too, feedback to the integrator can be implemented in conjunction with the differential element 52, so that its value cannot increase further, depending on the anti-windup strategy used, as soon as the calculated values ​​47 for the control variables are above or below the control variable limits. A selection element 62 (in particular a minimum element) is then used to select the feedback loop with the highest value, thereby limiting the integral component either by the limiting element 48 or by the signals at inputs 56 and 58. The in Fig. The control structure shown in section 3 is, for example, one of several parallel controllers in a successive control system. This allows the integral component of the parallel controllers in a successive control system to be limited (corresponding to the signals at inputs 56 and 58) when the controller is not active or is specifying the resulting manipulated variable. If the control structure shown is not part of a successive control system, selection element 62 can be omitted. In element 68, the control deviation 61, or the control deviation multiplied by the integral factor 66, is corrected or adjusted by the value determined by selection element 62. For example, with appropriately chosen signs, it is subtracted so that the value integrated in integrator element 70 becomes smaller in absolute value. The feedback loops shown here are particularly relevant when the hydraulic machine is operated as a pump. Should the hydraulic machine, or...If the constant unit is operated as a motor and used for regeneration purposes, the direction of rotation would have to be taken into account when selecting the feedback, at least in the speed controller, since negative speeds can also occur here.

[0043] To mathematically determine the manipulated variable and ascertain its value within the inverse system model, the derivation of the model equations will be demonstrated below using pressure control as an example. Based on the assumption that the constant-displacement pump is driven directly by the electric machine (without excluding a gearbox between the two components) and delivers a volumetric flow rate into a control volume (pump line), the following differential equations are suitable for describing these two effects:

[0044] The following applies to the pressure build-up (Equation 1): CH⋅p˙=Vg2π⋅ω−QLoad

[0045] The following applies to the increase in speed (Equation 2): J⋅ω˙=T−Vg2π⋅p−k⋅ω

[0046] Where: p is the pressure or pump pressure, C H the hydraulic capacity of the pump line, V g the swallowing volume of the constant pump, Q Load The total volume flow rate drawn by the hydraulic consumers, which is referred to below as the disturbance variable, J is the combined inertia of the electric machine and pump, T is the drive torque or torque to be applied by the electric machine (the actual manipulated variable), k is a speed-dependent friction component, and ω is the angular velocity (angular frequency) or rotational speed. A dot above a quantity, as usual, symbolizes its time derivative. If the first of these two formulas is rearranged for ω, differentiated once with respect to time, and substituted into the second formula, the following expression for T is obtained to describe the manipulated variable (Equation 3): T=2πJCHVgp¨+2πkCHVgp˙+Vg2πp+2πVgQ˙Load+2πkVgQLoad

[0047] This equation illustrates why the corresponding derivatives of the trajectory are also needed, especially to enable dynamic feedforward control during transient transitions. It is also apparent that the underlying equations contain simplifications. For example, leakage and the pump's friction effects are inadequately described, as these are typically represented as characteristic curves whose values ​​depend on the pressure and rotational speed. It is therefore proposed to derive the equations for determining the manipulated variable T from the pump's efficiency characteristic curves, which can also account for potential nonlinearities. This then results in (Equation 4): CH⋅p˙=Qeff,Pmp(p,ω)−QLoad and (Equation 5): J⋅ω˙=T−Teff,Pmp(p,ω)

[0048] To determine the rotational speed or the angular frequency ω, the first of these equations is solved for Q. eff,Pmp This is the effective pump flow rate minus pump leakage, rearranged. Since Q Load Since this value is usually unknown, it can also be set to zero, and the effectively required pump flow rate for the desired pressure increase is obtained as the product of the hydraulic capacity and the current value of the first time derivative of the trajectory. To obtain the rotational speed or the corresponding angular frequency, the characteristic curve Q must be used. eff,Pmp (p, ω) are inverted beforehand and implemented, for example, on the control unit. The resulting characteristic map ω(p, Q) eff,PmpBy specifying the desired pressure and the differential volume flow required to build it up, the speed difference can be determined to achieve the desired pressure increase. This speed difference can then be added to the measured actual speed, depending on the sign of the first derivative of the setpoint, and subsequently used to determine all further values ​​from the necessary characteristic curves. In this case, however, it is no longer a pure feedforward control, since the actual speed is used. As an alternative, it is suggested to sum the determined speed differences depending on the sign of the first derivative of the trajectory and thus obtain an approximate target speed or target angular frequency.

[0049] After rearranging the first equation to solve for ω, it is differentiated with respect to time and substituted into the second equation. The result for the drive torque of the electric machine, the desired manipulated variable, is the following equation (Equation 6): T=J∂Qeff,Pmp(p,ω)∂ω⋅(CH⋅p¨+Q˙Load−∂Qeff,Pmp(p,ω)∂p⋅p˙)+Teff,Pmp(p,ω)

[0050] The partial derivatives of the volumetric flow characteristic map with respect to angular frequency or pressure can be determined offline and implemented on the control unit. Furthermore, it is conceivable that instead of implementing the aforementioned measured characteristic maps and their inverted and / or derivative maps on the control unit, neural networks generated using appropriate data could be implemented. This could result in savings in memory requirements and / or processing time.

[0051] Fig. Figure 4 illustrates a filter structure with which trajectory values ​​36 and associated derivative values ​​38, 40 can be determined from setpoints 32 for the controlled state variable. The filter structure shown is an example of a state variable filter 34, as in Fig. 3 shown.

[0052] The state variable filter presented here is familiar to those skilled in the art and does not require a detailed explanation at this point. It is based on the following transfer function of two simple PT1 filters connected in series (Equation 7). pdes,fil=1(Tfil⋅s+1)2⋅pdes After the inverse transformation from Laplace space to time space, the following differential equation for the pressure setpoint can be determined (equation 8): p¨des,fil=1Tfil2⋅(pdes−pdes,fil−2⋅Tfil⋅p˙des,fil)

[0053] where p ges the target value specification for the pressure, which is e.g. by the target values ​​32 of the Fig. 3, Fig. 4 is given, p des,fil is the filtered target value specification, according to the trajectory, and T fil 80 is the time constant or filter constant of the PT1 filter.

[0054] The resulting differential equation is represented in the signal flow diagram of the Fig. Figure 4 illustrates the corresponding feedback loops of the states integrated by the integrator elements 84 and 86, and the inclusion of the filter constant 80 (T). filThe necessary second time derivative of the trajectory is generated, which is to be realized by the actual specification of the setpoint values ​​32. Using the exemplary limiting element 82, the value of the derivative can be limited both upwards and downwards. This can be used, for example, to adapt the generation of the trajectory, including the time derivatives necessary for dynamic feedforward control, to the actual prevailing limitations of the manipulated variable (here, the drive torque of the electric machine). This can be particularly advantageous when, due to excessive thermal loads on the inverter and / or electric machine, the drive torque to be delivered or received (during deceleration of the drive or in the case of regeneration) needs to be limited.To calculate the limitations of the second time derivative, the equation 3 above, rearranged for p̈, is used for the torque T in function module 90, for example. The input variables, besides the torque 92 or a torque that can be applied by the electric drive, are the corresponding states p and ṗ, as well as the (actually unknown) disturbance variables Q. Load and Q Load (Consumer total volume flow rate 94 and its time derivative 96). T represents the predetermined (dynamically, e.g., dependent on the temperature of the electric drive) limited drive torque, which is provided directly by the inverter's control unit. Here again, characteristic-map-based equations can be used to improve control performance. For this purpose, equation 6 can be rearranged to solve for p̈ and the previously presented input variables can be substituted.

[0055] In Fig. Figure 4 shows time profiles for the trajectory (trajectory values ​​36) and its second derivative (derivative values ​​40), as obtained by the illustrated filter for a jump in the target values ​​(setpoints 32). A setpoint profile 100 of the specified setpoints, a derivative profile 102 of the second time derivative of the trajectory, and a trajectory profile 104 of the trajectory are shown. It can be seen that the abrupt transition of the setpoints in the setpoint profile 100 is transformed into a flattened and increasing transition in the trajectory profile 104. The duration of this transition in the trajectory profile 104 depends in particular on the filter constant T. fildepending on which value can be chosen to be suitable for a given hydraulic system, e.g., sufficiently large so that the hydraulic system is technically capable of actually following the trajectory. In the derivative curve 102 of the second temporal derivative of the trajectory and in the trajectory curve 104 of the trajectory, the curves that would occur without the limiting element 82 are shown as dashed lines.

[0056] Fig. Figure 5 illustrates the use of a disturbance observer or observer 112. This allows the control performance to be further increased, especially in the transient transition behavior.

[0057] The disturbance variable consumer total volume flow Q Load is normally unmeasurable and therefore represents an unknown disturbance variable for feedforward control and regulation. The observer 112 used here determines the first time derivative Q. Load the disturbance variable.

[0058] In Fig. 5 is the observer 112 in connection with already in the Fig. 1 and Fig. The 3 elements shown are in Fig. 5 are essentially shown schematically. For a more detailed description of these elements, please refer to the description of the Fig. 1 and Fig. 3 referred. Accordingly Fig. Figure 1 shows a constant displacement hydraulic machine 2 (e.g., a constant-displacement pump) driven by an electric drive comprising an electric motor 4 and an inverter 6. The machine delivers the hydraulic fluid via a hydraulic line 12 (or pump channel) to a valve arrangement 8 or to at least one hydraulic consumer 10. A pump pressure sensor 16 measures the pump pressure, i.e., the pressure of the hydraulic fluid in the hydraulic line 12. Fig. Figure 3 shows a state variable filter 34, an inverse stretch model 46, and a compensation controller 75. The actual value 54 of the pressure (which serves as an example of the controlled state variable) is fed from the pump pressure sensor 16 to the control structure according to Fig. 3 is transmitted to determine the control deviation for the compensation controller 75 from this and from the trajectory values ​​36 of the trajectory. The actuating value 50 determined by the control structure is in turn used to control the inverter 6, whereby the actuating value 50, as shown, can be used directly or within the framework of a higher-level control system, e.g., a secondary control system, in which an ultimately used actuating value is selected from actuating values ​​determined by several parallel controllers (of which the actuating value 50 is one).

[0059] The disturbance variable observer 112 determines, for example, the disturbance variable Q̇ from the current measured values ​​of the pump pressure (actual value 54, measured by the pump pressure sensor 16) as well as the drive torque 116 (indirectly determinable from the phase current of the electric machine) and the drive speed 118 (both quantities are provided, for example, by the inverter control of the inverter) as follows. Load , based on equations 1 and 2 above.

[0060] Equation 1 is differentiated with respect to time, and then equation 2 is substituted into it. The result after Q Load rearranged (equation 9): Q˙Load=Vg2πJ(Tmeas−k⋅ωmeas−Vg2π⋅pmeas)−CH⋅p¨meas

[0061] These include: p meas the measured pump pressure, T meas the drive torque and ω determined and transmitted by the inverter control measThe angular frequency determined and transmitted by the inverter control. The derivatives of can be p meas Here, a suitable second-order state variable filter can be used to determine Q, so that the measured quantity does not need to be derived directly. Alternatively, a Luenberger observer or a Kalman filter can be used. Furthermore, the characteristic-map-based equations 4 and 5 can also be used to determine Q. Load to use. To compensate for noise and model errors, the observer's output signal can be further filtered with a DT1 element. This eliminates potential offsets without reducing the actual effect, especially during dynamic changes in the disturbance. The observed disturbance 120 (i.e., here Q̇) LoadThe calculated disturbance variable can be provided to the state variable filter 34 and / or the inverse plant model 46 of the controlled system. An additional possibility is to integrate the inverse model 46 and the disturbance observer 112 into a Kalman filter and then use this as the plant model. In this case, the calculated disturbance variable does not act via the actual compensation controller 75, but rather via the corresponding elements of the dynamic feedforward control.

[0062] Fig. Figure 6 illustrates a further embodiment to prevent overshoot during transient transitions with abrupt behavior or large setpoint change rates, such as in setpoint behavior 100 of the Fig. 3. To reduce the I-component. The first time derivative of the trajectory is used to limit the I-component.

[0063] Despite setpoint adjustment (by the state variable filter 34), abrupt setpoint changes can lead to control deviations, which can be amplified by the integral component and thus cause overshoot of the controlled state variable. To address this problem, the first time derivative of the trajectory, i.e., the derivative values ​​38, can be fed to elements 122 and 124. These can be implemented, for example, as parameterizable equations or parameterizable characteristic maps and, in this case, provide values ​​dependent on the time derivative of the trajectory for limiting the control deviation 61 calculated in the difference element 60. This limitation can be achieved via the limiting element 126.For example, if large rates of change of the setpoints prevail, the control deviation for the integral component of the controller can be limited to significantly smaller values, so that the control deviation amplified by the integral factor 66 does not lead to an excessive increase in the manipulated variable component by the integrator 70. The integrator's task of compensating primarily steady-state control deviations can still be fully achieved, since in these cases the first derivative of the trajectory assumes small values, even zero. Furthermore, it is conceivable that higher derivatives of the trajectory could also be used to limit the integral component. In another possible embodiment, no limiting element 126 is used, but the integral factor 66 is directly influenced.The I-factor 66 is therefore adjusted depending on the first derivative of the trajectory, whereby the I-factor 66 is smaller (in absolute value) when the first time derivative of the trajectory is larger (in absolute value). It is also conceivable that the influence on the I-factor 66 is determined by a characteristic map, which again takes the first time derivative of the trajectory as its input.

Claims

[1] Method for determining control values ​​of a manipulated variable for a pressure medium supply system of a hydraulic system with at least one hydraulic consumer (10), wherein the pressure medium supply system comprises a hydraulic machine (2) and an electric drive (4, 6) coupled thereto and is controlled by the manipulated variable, wherein an inverse plant model (47) is given for the hydraulic system, which uses values ​​of a state variable to be controlled of the hydraulic system and of time derivatives of the state variable up to a predetermined maximum order as input values ​​and maps these to an output value for the manipulated variable; the method comprising: - Determining (110) trajectory values ​​(36) of a trajectory for the time course of the state variable and derivative values ​​(38, 40) for time derivatives of the course of the state variable up to the maximum order from given setpoint values ​​(32) for the state variable; - Determining (120) compensation values ​​(73) with a compensation controller (75) from a control deviation (61) between the trajectory values ​​(36) and actual values ​​(54) of the state variable; and - Determining (130) of control values ​​(50) for the manipulated variable using the inverse path model (47) from the trajectory values ​​(36), the derivative values ​​(38, 40) and the compensation values ​​(73). [2] Method according to claim 1, wherein when determining (130) the control values ​​(50) for the control variable, the derivative values ​​(40) of the time derivative with the maximum order are changed by the compensation values ​​in order to determine modified derivative values ​​(44); wherein the trajectory values ​​(36), the derivative values ​​(38) of time derivatives with an order smaller than the maximum order and the modified derivative values ​​(44) are used as input values ​​of the inverse path model (47); wherein in particular the compensation values ​​(73) are added to the derivative values ​​(40) of the time derivative with the maximum order in order to determine the modified derivative values ​​(44). [3] Method according to one of the preceding claims, wherein when determining (110) the trajectory values ​​(36) and the derivative values ​​(38, 40) a filter (34) whose order is equal to the maximum order is used to determine the trajectory values ​​(36) and the derivative values ​​from the specified setpoint values ​​(32) for the state variable. [4] Method according to claim 3, wherein the filter (34) comprises or is designed as a number of PT1 filters connected in series, the number being equal to the maximum order. [5] Method according to one of the preceding claims, wherein one or more characteristic maps for the hydraulic machine (2) are used in the inverse system model (47); wherein in particular a volume flow characteristic map, which indicates the volume flow delivered by the hydraulic machine and which depends on a pressure and a speed of the hydraulic machine (2), and / or a torque characteristic map, which indicates the torque occurring at the hydraulic machine (2) and which depends on the pressure and the speed of the hydraulic machine (2), and / or one or more characteristic maps for the derivatives of the volume flow characteristic map and / or the torque characteristic map with respect to pressure or speed, and / or one or more inverted characteristic maps, which are inverses of the volume flow characteristic map or the torque characteristic map resolved with respect to pressure or speed, are used. [6] Method according to one of the preceding claims, wherein when determining (110) the trajectory values ​​(36) and the derivative values ​​(38, 40) the derivative values ​​for one of the temporal derivatives, in particular for the one with the maximum order, are restricted upwards and / or downwards by respective first limit values. [7] Method according to claim 6, wherein the first limit values ​​are varied depending on the trajectory values ​​(36) and / or the derivative values ​​that are not limited, and / or on the actual value of the manipulated variable and / or on a possible range of values ​​of the manipulated variable and / or on estimated values ​​of at least one disturbance variable determined by a disturbance estimation method, which is an unobserved or unobservable quantity in the hydraulic system; wherein the at least one disturbance variable includes in particular a total volume flow rate of the at least one consumer and / or its time derivative. [8] Method according to one of the preceding claims, wherein a disturbance estimation method is used to determine estimates for at least one disturbance variable, which is an unobserved or unobservable quantity in the hydraulic system; and wherein the estimates for the at least one disturbance variable are used when determining the trajectory values ​​and the derivative values ​​and / or in the inverse system model (47); wherein the at least one disturbance variable includes in particular a total volume flow rate of the at least one consumer and / or its time derivative. [9] Method according to claim 7 or 8, wherein the disturbance estimation method uses an observer (112), in particular a Luenberger observer, or an iterative estimation method, in particular a Kalman filter. [10] Method according to one of claims 7 to 9, wherein in the disturbance estimation method actual values ​​of the torque (116) applied by the electric drive (2, 4) and / or the rotational speed (118) of the electric drive are used or evaluated. [11] Method according to one of the preceding claims, wherein when determining (130) the control values ​​for the control variable, output values ​​(47) of the inverse system model (47) are restricted to determine the control values ​​(50). [12] Method according to claim 11, wherein the compensation controller (75) has an integral part; wherein difference values ​​between the output values ​​and the restricted control values ​​(50) are determined and used in a first anti-windup function of the integral part. [13] Method according to one of the preceding claims, wherein the compensation controller (75) has an integral part; wherein a second anti-windup function of the integral part depends on the derivative values ​​(38, 40), in particular the derivative values ​​(38) of the first time derivative. [14] Method according to claim 13, wherein the second anti-windup function limits the control deviation for the integral term downwards and / or upwards by respective second limit values, the second limit values ​​depending on the derivative values ​​(38, 40), in particular the derivative values ​​(38) of the first time derivative; and / or wherein the second anti-windup function changes an integral factor depending on the derivative values ​​(38, 40), in particular the derivative values ​​(38) of the first time derivative; and where, in particular, the second restriction values ​​and / or the integral factor are smaller in absolute value for derivative values ​​(38, 40) that are larger in absolute value, especially for derivative values ​​(38) of the first time derivative that are larger in absolute value. [15] Method according to one of the preceding claims, wherein the manipulated variable is a torque to be applied by the electric drive (4, 6) or a rotational speed of the electric drive. [16] Method according to one of the preceding claims, wherein the state variable is a pressure in the hydraulic system, in particular a pressure at a connection of the hydraulic machine (2) or a hydraulic line (12) connected to the connection or a hydraulic channel connected to the connection, a rotational speed of the hydraulic machine or a volume flow of pressure medium delivered by the hydraulic machine. [17] Computing unit (20, 22) comprising a processor configured to perform the method according to any of the preceding claims. [18] Hydraulic system comprising a pressure medium supply system with a hydraulic machine (2) and an electric drive (4, 6) coupled thereto and a computing unit according to claim 17, wherein an inverse system model (47) is given for the hydraulic system which uses values ​​of a state variable to be controlled of the hydraulic system and of time derivatives of the state variable up to a predetermined maximum order as input values ​​and maps these to an output value for the manipulated variable. [19] Computer program comprising instructions which, when the program is executed by the computing unit of the hydraulic system according to claim 18, cause it to execute the method according to claims 1 to 16. [20] Computer-readable data carrier on which the computer program according to claim 18 is stored.

Images