JP7567488B2 - State estimation system, control system, and state estimation method - Google Patents

Description

The present invention relates to a state estimation system, a control system, and a state estimation method.

There is a state estimation device that estimates other state variables of a controlled object based on the measured state variables of the controlled object (see, for example, Patent Document 1).

Japanese Patent Application Publication No. 7-121205

However, if the number of measured state variables in a controlled object is small, for example, the conditions of observability or detectability are not satisfied, and therefore the state estimation device generally cannot estimate other state variables of the controlled object. Therefore, in order for the state estimation device to estimate the state variables of the controlled object, it is necessary to increase the number of measured state variables. In such cases, the number of sensors to be installed in the controlled object increases, and the cost of the control system that controls the controlled object increases.

The present invention has been made in consideration of the above-mentioned problems in the conventional technology, and aims to provide a state estimation system that can estimate the state variables of a controlled object even when the number of state variables measured in the controlled object is small.

The state estimation system of the present invention, which solves the above-mentioned problems, includes a nominal plant that models a controlled object, a combiner that combines a first state variable that is measured among a plurality of state variables of the controlled object and a second state variable that is output from the nominal plant and is different from the first state variable, and a state estimator that estimates the plurality of state variables of the controlled object based on the first and second state variables combined by the combiner.

The control system of the present invention includes a nominal plant that models a controlled object, a synthesizer that synthesizes a first state variable that is measured among a plurality of state variables of the controlled object and a second state variable that is output from the nominal plant and is different from the first state variable, a state estimator that estimates the plurality of state variables of the controlled object based on the first and second state variables synthesized by the synthesizer, and a controller that controls the controlled object based on the difference between target values of the plurality of state variables and the plurality of state variables estimated by the state estimator.

The state estimation method of the present invention includes a step of combining a measured first state variable among a plurality of state variables of a controlled object and a second state variable that is output from a nominal plant that models the controlled object and is different from the first state variable, and a step of estimating the plurality of state variables of the controlled object based on the combined first and second state variables.

The present invention provides a state estimation system that can estimate the state variables of a controlled object even when the number of state variables measured in the controlled object is small.

FIG. 1 is a diagram for explaining an overview of a crane 10. FIG. 2 is a diagram showing an example of the configuration of a crane 10. FIG. 2 illustrates an example of a control system 30. FIG. 13 is a diagram for explaining state variables. FIG. 13 is a diagram showing an example of a simulation result of the control system 30. FIG. 13 is a diagram showing an example of a simulation result of the control system 30. FIG. 2 is a diagram illustrating an example of a control system 31.

The present specification and accompanying drawings make clear at least the following:

<<<<Overview of the crane 10>>>
1 is a diagram showing an overview of an overhead crane 10 (hereinafter simply referred to as a crane) capable of movement along three axes: traveling (x-axis), traversing (y-axis), and hoisting (z-axis). The crane 10 includes a trolley 20 that is entirely movable in the x-axis direction and that is movable in the y-axis direction along a girder (not shown), and a load 21 suspended by a wire rope that can be hoisted up or down from the trolley 20.

＜＜＜トロリー２０のモデリング＞＞＞ Hereinafter, in this embodiment, the coordinates of the trolley 20 are expressed as (x, y, 0), and the coordinates of the suspended load 21 are expressed as (x ^~ , y ^~ , z ^~ ). Note that "x ^~ ,""y ^~ ," and "z ^~ " are the same as the symbols (1) to (3) below. Also, the swing width θ of the suspended load 21 is evaluated by the angle φ of the suspended load 21 projected onto the xy plane, the angle α of the suspended load 21 projected onto the xz plane, and the angle β of the suspended load 21 projected onto the yz plane.

<<<Modeling of Trolley 20>>>

Figure 2 shows an example of the configuration of the trolley 20 and the suspended load 21. For convenience, only the minimum configuration required for modeling the trolley 20 and the suspended load 21 is shown.

The trolley 20 includes a control system 30, a motor 40, and a sensor 41. Although details will be described later, the control system 30 outputs an x-axis command value u _x , a y-axis command value u _y , and a z-axis command value u _z to the motor 40. Note that hereinafter, the command values for the three axes, x-axis to y-axis, will be collectively referred to as command value u.

The motor 40 moves the crane 10 itself in the x-axis direction based on the command value u _x . The motor 40 also moves the trolley 20 in the y-axis direction along the girder (not shown) by rotating the wheels of the trolley 20 based on the command value u _y . Furthermore, the motor 40 winds up or lowers the wire rope to which the load 21 is connected in the z-axis direction based on the command value u _z .

When the trolley 20 operates according to a command value u from the control system 30, the following formulas (1) to (3) are established. In the following, in this embodiment, the "movement of the trolley 20" includes the movement of the crane 10 itself and the movement of the trolley 20 on the girder.

Here, the crane 10 and trolley 20 have first-order lag characteristics, the gain of the motor 40 is "K", and the time constant of the motor 40 is "T". Therefore, for example, the gain of the motor 40 in the x-axis direction can be expressed as Kx, and the time constant of the motor 40 in the x-axis direction can be expressed as Tx. Also, the term on the left side of equation (1) is an expression of "x" differentiated twice with respect to time t, and the symbol for x on the right side is an expression of "x" differentiated once with respect to time t.

<<<Modeling of Suspended Load 21>>>
By transforming the equation of motion for the suspended load 21 and assuming that the mass of the suspended load 21 is “m”, the gravitational acceleration is “g”, the damping coefficient in the x-direction is “D _α ”, and the damping coefficient in the x-direction is “D _β ”, the following equations (4) and (5) hold for the suspended load 21.

<<<<State Space Model>>>
Since the formulas (1) to (5) include second-order differentials, they are converted here into expressions including only first-order differentials in order to facilitate the analysis of the crane 10. Specifically, the formulas (1) to (5) are expressed as the following formula (6).

Here, "x ₁ ", "y ₁ ", "z ₁ ", "x ^∼ ₁ ", and "y ^∼ ₁ " are symbols defined to avoid symbols for second-order differentiation. Furthermore, by rearranging equation (6) into a state space model, the following equation (7) is obtained.

In equation (7), the state vector xp including 10 state variables in the state space model, the command value vector u, the matrix A, and the matrix B are as shown below.

In this way, the crane 10 controlled by the control system 30 can be represented by a state space model such as that shown in equation (7).

Here, of the 10 state variables, x is the position of the trolley 20 in the x-axis direction (traveling), dx/dt is the speed of the trolley 20 in the x-axis direction, Δx is the swing width of the load 21 in the x-axis direction, and dΔx/dt is the speed of the swing width of the load 21. Furthermore, y is the position of the trolley 20 in the y-axis direction (traversal), dy/dt is the speed of the trolley 20 in the y-axis direction, Δy is the swing width of the load 21 in the y-axis direction, and dΔy/dt is the speed of the swing width of the load 21. Furthermore, z is the hoisting position, and dz/dt is the hoisting speed.

The sensor 41 in FIG. 2 of this embodiment measures only the hoisting speed (dz/dt) out of the 10 state variables. In other words, in the crane 10, only one state variable (dz/dt) out of the 10 state variables is measured. The control system 30 of this embodiment controls the crane 10 based on only the measured one state variable (dz/dt), as will be described in detail later. The state variable (dz/dt) indicating the hoisting speed of the load 21 corresponds to the "first state variable."

<<<<Details of Control System 30>>>
3 is a block diagram showing an example of the configuration of the control system 30. The control system 30 is a discrete-time system that controls the crane 10, and includes controllers 50, 57, adders 51, 54, 56, a nominal plant 52, a combiner 53, and a state estimator 55. First, a block diagram of the crane 10, which is the object to be controlled, will be described. In the control system 30 of this embodiment, the nominal plant 52, the combiner 53, and the state estimator 55 correspond to a "state estimation system."

<<About the crane 10>>
The state equation of the crane 10 is expressed by the above-mentioned equation (7), and further, the output equation of the crane 10 is expressed by equation (8).

The crane 10 is composed of blocks 100, 102 to 104 that embody equations (7) and (8), and an adder 101. Block 100 is a block that multiplies the input command value by matrix B of equation (7), and adder 101 adds the output from block 100 and the output from block 103. Block 102 also integrates the output from adder 101 and outputs state vector xp. Block 103 is a block that multiplies state vector xp by matrix A of equation (7). In this way, blocks 100, 102, 103, and adder 101 calculate equation (7) for a command value (e.g., command value u) input to the crane 10.

And block 104 is a block corresponding to matrix C in equation (8) above. In this embodiment, as described above, sensor 41 of crane 10 measures only the hoisting speed (dz/dt). Therefore, block 104 outputs only the hoisting speed (dz/dt) of the 10 state variables as output yp.

<<Regarding the Controller 50>>
The controller 50 outputs a command value u for controlling the crane 10 based on a command value u _FF from a controller (not shown) of the crane 10 and a state vector x^p estimated by a state estimator 55 described later. Note that the command value u _FF in this embodiment is a command value for so-called feedforward control of the crane 10. The controller 50 is configured to include a nominal plant 60 (G _FF ), adders 61 and 63, and a block 62.

The nominal plant 60 is a block that models the crane 10. In this embodiment, the "nominal plant" is a so-called real plant simulator that models a controlled object, and is also called a nominal model. For this reason, the nominal plant includes a block (not shown) that executes equations (7) and (8) in the same way as the crane 10. When a command value _uFF is input, the nominal plant 60 in this embodiment outputs a state vector xref including ten state variables based on equations (7) and (8). The state vector xref corresponds to a "target value of the state variables" when controlling the crane 10.

An adder 61 calculates the difference e between the state vector xref output from the nominal plant 60 and the state vector x^p estimated by the state estimator 55. A block 62 (K _FB ) is a block in which a gain, a first-order lag element, and the like for feedback controlling the crane 10 are set. Therefore, the block 62 outputs a command value uk for feedback controlling the crane 10 based on the difference e and the set gain, etc. Then, an adder 63 adds a command value u _FF for feedforward control of the crane 10 and a command value uk for feedback control, and outputs the result as a command value u.

If there is no disturbance or noise (described later) in the control system 30 and the nominal plant 60 and each block accurately simulates the crane 10, the state vector xref and the estimated state vector x^p will be equal, and the difference e will be zero. In such a case, the command value uk will also be zero, and the crane 10 will be controlled only by feedforward control. However, since the actual control system 30 is affected by disturbances and noise, it is difficult to accurately model the crane 10 within each block such as the nominal plant 60. However, since the control system 30 also performs feedback control on the crane 10, it is possible to operate the crane 10 more stably.

<<Adder 51>>
The adder 51 adds a command value u from the controller 50, a command value uL from a controller 57(L) described later, a disturbance d, and noise w. The command value uL from the controller 57 is a command value mainly for canceling the disturbance d and the noise w.

<<About Nominal Plant 52 (Gsim)>>
The nominal plant 52 is a block that models the crane 10, similar to the nominal plant 60. When the command value from the adder 51 is input, the nominal plant 52 outputs four state variables as an output yps based on the formulas (7) and (8). Here, the four state variables output by the nominal plant 52 are the state variables x, dx/dt, y, and dy/dt, as shown in FIG. 4. Note that FIG. 4 is a diagram for explaining the details of the state variables. Here, the "acquisition method" written in the first row and fourth column of the table in FIG. 4 indicates how the state variables are acquired in the control system 30. The state variables x, dx/dt, y, and dy/dt are calculated by the nominal plant 52, and therefore are written as "calculated".

Here, the state variables x, dx/dt, y, and dy/dt correspond to the "second state variables", and the state variables x and dx/dt correspond to the "variables indicating the state of the trolley on the first axis (x-axis)". Furthermore, the state variables y and dy/dt correspond to the "variables indicating the state of the trolley on the second axis (y-axis)". Therefore, the nominal plant 52 corresponds to the "nominal plant that outputs the second state variables".

<<About the combiner 53>>
The synthesizer 53 outputs a state vector ypt used for state estimation based on the output yp from the crane 10 and the output yps from the nominal plant 52. Here, the synthesizer 53 includes an integrator (not shown) that integrates the output yp (dz/dt). Specifically, the integrator integrates the hoisting speed (dz/dt) of the load 21 to calculate the hoisting position z of the load 21. Then, the synthesizer 53 synthesizes the hoisting position z of the load 21, the hoisting speed (dz/dt), and the output yps, and outputs the state vector ypt. Therefore, the synthesizer 53 outputs a state vector ypt including six state variables out of the ten state variables.
ypt=(x, dx/dt, 0, 0, y, dy/dt, 0, 0, z, dz/dt) ^T

In this embodiment, the combiner 53 includes an integrator (not shown) that integrates the hoisting speed and calculates the hoisting position, but this is not limited to this. The integrator may be provided between the crane 10 and the combiner 53. Also, the combiner 53 may directly combine the output yp from the crane 10 and the output yps from the nominal plant 52.

<<Adder 54>>
The adder 54 is a block that adds the state vector ypt and noise v of the sensor 41, etc. The adder 54 is a block that is drawn for convenience in the block diagram of the control system 30 in order to take into account the noise v.

<<State Estimator 55>>
The state estimator 55 is a Kalman filter type estimator that estimates ten state variables based on the command value u from the controller 50 and the state vector ypt. The state estimator 55 includes a nominal plant 70, an adder 71, and a block 72. The nominal plant 70, like the other nominal plants and the crane 10, includes a block (not shown) that executes equations (7) and (8). Details of the nominal plant 70 will be described later.

<<Adder 71 and Block 72 (K _K )>>
An adder 71 calculates the difference ε^y between the output ypt+v from the adder 54 and the output y^p from the nominal plant 70. A block 72 multiplies the difference ε^y by the gain of the Kalman filter. In this embodiment, the gain of the Kalman filter is appropriately adjusted so that the difference ε^y becomes small.

<<Details of Nominal Plant 70>>
The nominal plant 70 includes blocks 80, 82 to 84, and an adder 81. Here, the blocks 80, 82 to 84 are similar to the blocks 100, 102 to 104, respectively. The output ypt included in the adder 54 includes six of the ten state variables. Therefore, the matrix C is set so that the block 84 also outputs the same six state variables as the output ypt. The adder 81 is similar to the adder 101 except that it adds the output from the block 72.

By using the state estimator 55 configured in this way, the block 82 can estimate the state vector x^p including 10 state variables. In other words, as shown in FIG. 4, of the 10 state variables, four state variables Δx, dΔx/dt, Δy, and dΔy/dt are "state estimated." In addition, the state vector x^p is output to the controller 50 described above and the adder 56 described later.

<<Adder 56 and Controller 57>>
The adder 56 and the controller 57 are blocks for removing, for example, disturbance d and noise w. Since the actual crane 10 is provided with only the sensor 41, only one state variable (dz/dt) is output as described above. Therefore, although not shown for convenience, the control system 30 is provided with at least blocks 100, 102, and 103 that simulate the crane 10, and an adder 101, in addition to the actual crane 10. As a result, the state vector xp is output from the block 102.

The adder 56 calculates the difference ε between the state vector xp from the block 102 and the estimated state vector x^p. The controller 57 (L) outputs a command value uL to the adder 51 so that the difference ε becomes smaller. As a result, disturbances d etc. included in the command value output from the adder 51 to the actual crane 10 and the nominal model simulating the crane 10 (a model including the block 100 etc.) are reduced.

<<<<Simulation results>>>
5 is a diagram showing an example of a simulation result of the control system 30. Here, it is assumed that in the control system 30, feedforward control based on the command value u _FF , feedback control based on the command value uk, and disturbance feedback (DFC) based on the command value uL are functioning. In addition, the command value u _FF is input so that the trolley 20 accelerates from 0 to 10 seconds in each of the x-axis and y-axis directions, and moves at a constant speed after 10 seconds. Furthermore, a disturbance d is input at 15 seconds in the x-axis direction and at 20 seconds in the y-axis direction.

The upper part of Figure 5 shows a comparison between the estimated result (solid line) of the swing width Δx (here, written as θx) of the suspended load 21 in the x-axis direction and the output result (dotted line) from the crane 10. The lower part of Figure 5 shows a comparison between the estimated result (solid line) of the swing width Δy (here, written as θy) of the suspended load 21 in the y-axis direction and the output result (dotted line) from the crane 10.

In this way, θx of the estimated state vector x^p and θx of the state vector xp from the crane 10 are almost the same. Therefore, as shown in FIG. 4, the state estimator 55 of this embodiment can accurately estimate the state variables of the crane 10 even if there is only one output actually measured from the sensor 41.

6 is a diagram showing another example of the simulation result of the control system 30. Here, it is assumed that only the feedforward control based on the command value _uFF and the feedback control based on the command value uk function in the control system 30. Therefore, the simulation result is for the control system 30 of FIG. 3 excluding the adder 56 and the controller 57.

The upper part of Fig. 6 shows a comparison between the estimated result (solid line) of the swing width Δx (here, written as θx) of the suspended load 21 in the x-axis direction and the output result (dotted line) from the crane 10. The lower part of Fig. 6 shows a comparison between the estimated result (solid line) of the swing width Δy (here, written as θy) of the suspended load 21 in the y-axis direction and the output result (dotted line) from the crane 10.

Compared to Figure 5, the difference between the estimated results and the output results from the crane 10 for both θx and θy is slightly larger, but the two match well. Therefore, even if disturbance feedback control (DFC) is not provided in the control system 30, the state estimator 55 can accurately estimate the state variables of the crane 10.

Control System 31
FIG. 7 is a diagram showing an example of the control system 31. Unlike the control system 30 which operates discretely, the control system 31 is a system which operates continuously. In FIG. 3 and FIG. 7, blocks with the same reference numerals are the same except that they operate discretely or continuously. In the control system 31, a Luenberger type state estimator 58 is used instead of the Kalman filter type state estimator 55. For this reason, the state estimator 58 will be mainly described here. In FIG. 7, the block 102 of the crane 10 and the block 82 of the nominal plant 70 also operate continuously, so the notation of these blocks has been changed from "Z ^-1 " to "∫". In the control system 31, the nominal plant 52, the combiner 53, and the state estimator 58 correspond to a "state estimation system".

<<State Estimator 58>>
The state estimator 58 includes a nominal plant 70, an adder 71, and a block 75. Here, the nominal plant 70 and the adder 71 are the same as the blocks in Fig. 3 except that they operate continuously, so detailed description thereof will be omitted.

Block 75(h) is a block that multiplies the difference ε^y by a predetermined gain. In this embodiment, the predetermined gain is adjusted so that the difference ε^y is small. Therefore, the state estimator 58 can accurately match the state vector x^p, which is the estimation result, with the state vector xp from the crane 10. Even when such a state estimator 58 is used, the control system 31 operates in the same manner as the control system 30. In other words, even when there is only one state variable measured by the sensor 41, the state estimator 58 of the control system 31 can accurately estimate the four state variables as shown in FIG. 4 by using the four state variables from the nominal plant 52.

<<<Others>>>
In this embodiment, the crane 10 has been described as an example of a controlled object, but the controlled object is not limited to this. For example, the controlled object may be a refrigeration system that keeps the temperature inside the storage constant. In a refrigeration system, a method of controlling the superheat temperature at the evaporator outlet to be constant (constant superheat control) and a method of controlling the suction pressure of the compressor to be constant (constant suction pressure control) are generally used.

However, since pressure sensors are expensive, costs can be reduced if state estimation can be performed instead of using sensors. On the other hand, thermocouples that measure superheat temperature are inexpensive, and since the characteristics of superheat vary greatly depending on operating conditions, it is desirable to use a sensor. Thus, in a refrigeration system, an embodiment can be considered in which the suction pressure is estimated using only a thermocouple sensor and a refrigeration system model.

In such a case, the output from the thermocouple sensor corresponds to the "first state variable," the refrigeration system model corresponds to the "nominal plant," and the value indicating the pressure of the refrigeration system model corresponds to the "second state variable."

== ...
The control systems 30 and 31 according to an embodiment of the present invention have been described above. In the control systems 30 and 31, only one state variable, the hoisting speed (dz/dt), among the ten state variables is measured by the sensor 41, and the hoisting position (z) is calculated from the measurement result. The synthesizer 53 synthesizes these two state variables with the four state variables (x, dx/dt, y, dx/dt) output from the nominal plant 52. Then, as shown in FIG. 4, the state estimator 55 can accurately estimate the remaining four state variables (Δx, dΔx/dt, Δy, dΔy/dt,) based on these six state variables (for example, FIG. 5). Therefore, in this embodiment, even if the number of sensors is small, the state variables of the crane 10 to be controlled can be estimated.

In addition, as a state estimator, for example, a Kalman filter type state estimator 55 or a Luenberger type state estimator 58 can be used.

The crane 10, which is the object to be controlled, includes a suspended load 21 that is susceptible to characteristic fluctuations and disturbances, and a trolley 20 that is not susceptible to characteristic fluctuations and disturbances. Note that "characteristic fluctuations" here include, for example, fluctuations in the suspended load 21 itself, which changes in weight and shape. Furthermore, "disturbances" include, for example, noise and changes in the surrounding environment (wind, temperature, humidity) when the crane 10 is operated. For this reason, "susceptible to characteristic fluctuations and disturbances" refers to, for example, large modeling errors (or model fluctuations) when an actual object is modeled.

In this embodiment, the state variables of the trolley 20 and the suspended load 21, which is susceptible to characteristic fluctuations and disturbances, are measured by the sensor 41, and the state variables of the trolley 20, which is less susceptible to characteristic fluctuations and disturbances, use the output from the nominal plant 52. In this way, when there are multiple state variables, by installing the sensor 41 for the object that is susceptible to characteristic fluctuations and disturbances and acquiring the state variables, all state variables can be estimated with high accuracy.

As described above, in this embodiment, the element of the crane 10 that is susceptible to characteristic fluctuations and disturbances is the suspended load 21 (first element), and the element that is less susceptible to characteristic fluctuations and disturbances is the trolley 20 (second element).

The nominal plant 52 also outputs to the combiner 53 the state variables of the trolley 20 on the x-axis (e.g., position x) and the state variables of the trolley 20 on the y-axis (e.g., position y) on the xy plane. In this case, the state estimator 55 can estimate the state variables with higher accuracy than when the nominal plant 52 outputs the state variables of only one of the two axes.

In addition, in this embodiment, the sensor 41 measures only the hoisting speed (dz/dt) out of the 10 state variables. A specific block of the control system 30 can determine the hoisting position (z) of the load by integrating the hoisting speed. Therefore, in this embodiment, the number of state variables to be measured can be reduced.

In addition, in the control systems 30 and 31 of this embodiment, the controller 50 outputs a command value u that executes feedback control and feedforward control based on the difference between the state vector xref, which indicates the target value of the state variable, and the estimated state vector x^p. As a result, the crane 10 is stably controlled.

The calculations (steps) performed by the synthesizer 53 and the state estimator 55 may be performed by a computer such as a DSP (digital signal processing circuit). In such a case, the DSP executes the state estimation method executed by the control system 30.

In this embodiment, the sensor 41 acquires only the winding speed (dz/dt), but it may also acquire only the winding position (z), for example. In addition, the nominal plant 52 outputs four state variables, but it may output a smaller number of state variables if the conditions of observability or detectability are satisfied. Furthermore, the state variables output by the nominal plant 52 may be any state variables other than the measured state variables.

The above embodiments are intended to facilitate understanding of the present invention, and are not intended to limit the scope of the present invention. Furthermore, the present invention may be modified or improved without departing from the spirit of the present invention, and it goes without saying that the present invention includes equivalents.

10 Crane 20 Trolley 21 Suspended load 30, 31 Control system 40 Motor 41 Sensor 50, 57 Controller 51, 54, 56, 61, 63, 71, 81, 101 Adder 53 Synthesizer 52, 60, 70 Nominal plant 55, 58 State estimator 62, 72, 75, 80, 82, 83, 84, 100, 102, 103, 104 Block

Claims (8)

A nominal plant that models the controlled object,
a synthesizer that synthesizes a first state variable that is measured among the plurality of state variables of the controlled object and a second state variable that is output from the nominal plant and is different from the first state variable;
a state estimator that estimates the plurality of state variables of the controlled object based on the first and second state variables combined by the combiner;
A state estimation system comprising: The state estimation system according to claim 1 ,
The state estimator is a Kalman filter type or a Luenberger type state estimator.
State estimation system. 3. A state estimation system according to claim 1, further comprising:
the controlled object includes a first element and a second element that is less affected by characteristic fluctuations or disturbances than the first element,
the first state variable is a variable indicating a state of the first element,
The second state variable is a variable indicating a state of the second element.
State estimation system. The state estimation system according to claim 3,
The control object is a crane including a trolley movable on a predetermined plane and a suspended load,
The first element is the suspended load,
The second element is the trolley.
State estimation system. The state estimation system according to claim 4,
The second state variables include a variable indicating a state of the trolley in a first axis and a variable indicating a state of the trolley in a second axis among two different axes on the plane,
State estimation system. The state estimation system according to claim 4 or 5,
The first state variable indicates a hoisting speed of the load.
State estimation system. A state estimation system according to any one of claims 1 to 6,
a controller that controls the control target based on a difference between target values of the plurality of state variables and the plurality of state variables estimated by the state estimator;
A control system comprising: A step of combining a first state variable that is measured among a plurality of state variables of a controlled object and a second state variable that is output from a nominal plant that models the controlled object and is different from the first state variable;
estimating the plurality of state variables of the controlled object based on the combined first and second state variables;
A state estimation method comprising:

Images