Orbit prediction method and control system

Abstract

In a method for predicting and controlling the trajectory of a controlled object, the aim is to maintain high prediction accuracy while enabling rapid recovery from sudden disturbances. [Solution] For the controlled object, a predicted trajectory is calculated according to a first trajectory calculation formula based on the autoregressive method (S12), and the object is moved along the calculated predicted trajectory (S13). The position of the moved controlled object is detected (S15) and compared with the predicted trajectory (S16). If the error exceeds a threshold, it is determined to be an anomaly, and a predicted trajectory is calculated according to a second trajectory calculation formula which has fewer terms than the first trajectory calculation formula (S19), and the controlled object is moved along the calculated predicted trajectory (S1A).

Description

Technical Field

[0001] This disclosure relates to a technique for predicting the trajectory of a controlled object and controlling its movement.

Background Art

[0002] In control devices such as semiconductor chip mounters, for example, when vibrations occur during the operation and stop of a head, this vibration can cause a deterioration in position accuracy or a loss of manufacturing time waiting for the residual vibration to subside.

[0003] Patent Document 1 discloses the configuration of a vibration diagnosis device for a rotating machine. In this configuration, the coefficients and order of an autoregressive model are determined from the latest vibration data of the rotating machine, a vibration prediction value is calculated from the determined autoregressive model, a residual between the vibration prediction value and the measured vibration value is calculated, and the magnitude of the noise component of the autoregressive model is calculated. By comparing the residual, the magnitude of the noise component, and a threshold value, a mutation event or a noise event in the vibration data is determined.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] In control devices, there is a method of using a high-speed camera to grasp the position of a controlled object in a short time and operate it so as not to be affected by vibrations or the like. However, in this method, there is a delay from imaging to image processing and control until the device operates, so the control device cannot respond instantaneously to vibrations.

[0006] Also, there is a feedforward method of operating in consideration of the delay in advance, but it cannot be applied when the operation is not determined depending on the situation such as vibrations.

[0007] Furthermore, it is difficult to respond to sudden, unexpected shocks and vibrations that are not anticipated, requiring a rapid response that can be achieved instantaneously.

[0008] The purpose of this disclosure is to provide a method and control system that can perform orbit prediction that allows for rapid recovery from sudden disturbances while maintaining high prediction accuracy. [Means for solving the problem]

[0009] This disclosure provides a method for predicting the trajectory of a controlled object using an autoregressive method and moving the controlled object according to the predicted trajectory, comprising: (a) calculating the predicted trajectory of the controlled object according to a first trajectory calculation formula and moving the controlled object along the calculated predicted trajectory; (b) detecting the position of the moved controlled object; (c) comparing the predicted trajectory calculated in step (a) with the position of the controlled object detected in step (b), determining that the error between the predicted trajectory and the detected position is normal if it does not exceed a threshold, and abnormal if it exceeds a threshold; (d) moving the controlled object along the predicted trajectory according to the first trajectory calculation formula when it is determined to be normal in step (c); and (e) calculating the predicted trajectory according to a second trajectory calculation formula having fewer terms than the first trajectory calculation formula and moving the controlled object along the calculated predicted trajectory when it is determined to be abnormal in step (c). [Effects of the Invention]

[0010] This disclosure makes it possible to predict and control the trajectory of a controlled object while maintaining high prediction accuracy and enabling rapid recovery from sudden disturbances. [Brief explanation of the drawing]

[0011] [Figure 1] A schematic configuration example of a semiconductor chip mounting machine, which is an example of a control device to which the trajectory prediction method described herein can be applied. [Figure 2](a) and (b) are images of trajectory prediction in response to vibrations. [Figure 3] Diagram to explain autoregression [Figure 4] Trajectory changes and sudden disturbance models when disturbances occur [Figure 5] Relationship between the number of AR terms, error rate, and AR return time [Figure 6] Example configuration of a control system according to the embodiment [Figure 7] Graph showing the effects of the embodiment [Figure 8] Flowchart showing an example of a trajectory prediction method according to the embodiment [Figure 9] Flowchart showing an example of recovery procedures from an abnormal situation [Figure 10] An example of a user interface [Figure 11] Diagram illustrating model predictive control [Modes for carrying out the invention]

[0012] (overview) A trajectory prediction method according to an aspect of the present disclosure is a method for predicting the trajectory of a controlled object using an autoregressive method and moving the controlled object according to the predicted trajectory, comprising: (a) calculating the predicted trajectory of the controlled object according to a first trajectory calculation formula and moving the controlled object along the calculated predicted trajectory; (b) detecting the position of the moved controlled object; (c) comparing the predicted trajectory calculated in step (a) with the position of the controlled object detected in step (b), determining that the error between the predicted trajectory and the detected position is normal if it does not exceed a threshold, and determining that it is abnormal if it exceeds a threshold; (d) moving the controlled object along the predicted trajectory according to the first trajectory calculation formula when it is determined to be normal in step (c); and (e) calculating the predicted trajectory according to a second trajectory calculation formula having fewer terms than the first trajectory calculation formula and moving the controlled object along the calculated predicted trajectory when it is determined to be abnormal in step (c).

[0013] According to this configuration, for the controlled object, a predicted trajectory is calculated according to the first trajectory calculation formula based on the autoregressive method, and the controlled object is moved along the calculated predicted trajectory. The position of the moved controlled object is detected and compared with the predicted trajectory. If the error exceeds the threshold value, it is determined as an abnormality, and a predicted trajectory is calculated according to the second trajectory calculation formula with fewer terms than the first trajectory calculation formula, and the controlled object is moved along the calculated predicted trajectory. Thereby, when a sudden disturbance occurs and the actual trajectory of the controlled object deviates greatly from the prediction, the trajectory prediction is performed using a trajectory calculation formula with fewer terms, where the return time until the predicted trajectory returns to the actual trajectory is shorter. Therefore, it is possible to perform a trajectory prediction that can quickly return to a sudden disturbance while maintaining high prediction accuracy.

[0014] In the trajectory prediction method according to the above aspect, after step (e), steps (b) and (c) are executed again. When it is determined to be normal in step (c) executed again, the predicted trajectory is calculated according to the first trajectory calculation formula or the third trajectory calculation formula with more terms than the second trajectory calculation formula, and the controlled object is moved along the calculated predicted trajectory. This is also possible.

[0015] Thereby, when a sudden disturbance occurs and the actual trajectory of the controlled object deviates greatly from the prediction, the predicted trajectory can be more stably returned to the actual trajectory.

[0016] In the trajectory prediction method according to the above aspect, before the movement of the controlled object, a step (f) of preparing a plurality of trajectory calculation formulas is provided. In step (e), the trajectory calculation formula with the smallest error between the predicted trajectory and the detected position of the controlled object is selected from the plurality of trajectory calculation formulas prepared in step (f) as the second trajectory calculation formula. This is also possible.

[0017] Thereby, when a sudden disturbance occurs and the actual trajectory of the controlled object deviates greatly from the prediction, the predicted trajectory can be more quickly returned to the actual trajectory.

[0018] In the trajectory prediction method according to the above embodiment, the method may include a step in which, before the movement of the controlled object, the user sets the number of terms and coefficients of the trajectory calculation formula, as well as the threshold, via a user interface.

[0019] In the trajectory prediction method according to the above embodiment, step (b) may be performed based on an image captured by an imaging device.

[0020] The control system according to this disclosure predicts the trajectory of a controlled object using an autoregressive method and moves the controlled object according to the predicted trajectory, and comprises a detection unit for detecting the position of the controlled object, a calculation unit for calculating the predicted trajectory of the controlled object according to a trajectory calculation formula, a control unit for controlling the movement of the controlled object along the predicted trajectory calculated by the calculation unit, and a disturbance determination unit for comparing the predicted trajectory calculated by the calculation unit with a detected position detected by the detection unit, and determining that a disturbance has occurred if the error between the predicted trajectory and the detected position exceeds a threshold, wherein the calculation unit calculates the predicted trajectory according to a first trajectory calculation formula, and when the disturbance determination unit determines that a disturbance has occurred, it calculates the predicted trajectory according to a second trajectory calculation formula which has fewer terms than the first trajectory calculation formula.

[0021] In this configuration, the calculation unit calculates the predicted trajectory of the controlled object according to a trajectory calculation formula based on autoregression, and the control unit moves the controlled object along the calculated predicted trajectory. The detection unit detects the position of the moved controlled object, and the disturbance determination unit compares the predicted trajectory with the position of the controlled object. If the error exceeds a threshold, it determines that a disturbance has occurred. When it is determined that a disturbance has occurred, the calculation unit calculates the predicted trajectory according to a second trajectory calculation formula, which has fewer terms than the first trajectory calculation formula. The control unit moves the controlled object along the calculated predicted trajectory. As a result, when a sudden disturbance occurs and the actual trajectory of the controlled object deviates significantly from the prediction, trajectory prediction is performed using a trajectory calculation formula with fewer terms, which shortens the recovery time until the predicted trajectory returns to the actual trajectory. Therefore, trajectory prediction that can quickly recover from sudden disturbances can be performed while maintaining high prediction accuracy.

[0022] In the control system according to the above embodiment, the system may be provided with a user interface in which the user sets the number of terms and coefficients of the trajectory calculation formula and the threshold, and when the disturbance determination unit determines that a disturbance has occurred, the control system outputs, via the user interface, a statement indicating that a disturbance has been determined to have occurred and the trajectory calculation formula used to calculate the predicted trajectory.

[0023] (Embodiment) The embodiments will be described in detail below with reference to the drawings.

[0024] The embodiments described below are all comprehensive or specific examples. The numerical values, shapes, materials, components, arrangement and connection configurations of components, steps, and the order of steps shown in the following embodiments are examples only and are not intended to limit this disclosure. Furthermore, among the components in the following embodiments, those not described in the independent claim representing the highest-level concept will be described as optional components.

[0025] Figure 1 shows a schematic example of a semiconductor chip mounting machine. The semiconductor chip mounting machine shown in Figure 1 is an example of a control device to which the trajectory prediction method according to this disclosure can be applied. In the configuration of Figure 1, the head 11 holds the semiconductor chip 12 and mounts the semiconductor chip 12 at a predetermined position on the substrate 14 placed on the stage 13. The head 11 can move and stop in the vertical and horizontal directions. A camera 15 captures, for example, alignment marks attached to the substrate 14, and controls the position of the head 11 according to the position recognized from the captured image.

[0026] Here, if vibration occurs in the head 11, it is necessary to move the head 11 along a trajectory that follows that vibration. However, as shown in Figure 2(a), the movement of the head 11 may lag behind the vibration due to, for example, image processing delay, control delay, response delay, etc. To solve this problem, as shown in Figure 2(b), it is necessary to predict the trajectory in response to the vibration and move the head 11 to the predicted position. And it is preferable to use a calculation using a simple formula so that the predicted position can be calculated immediately. For this reason, an autoregressive method (AR, ARIMA) will be used.

[0027] Figure 3 is a diagram illustrating the autoregressive method. As shown in Figure 3, the autoregressive method calculates the predicted future trajectory by multiplying past actual trajectory data by coefficients and adding them together. Figure 3 shows an example where the predicted trajectory is calculated using actual trajectory data from the last three trajectories. Here, the number of actual data points used for trajectory prediction is called the "AR term count." In the example in Figure 3, the AR term count is 3. Furthermore, each coefficient in the formula used for the calculation can be determined in advance by learning from statistical probabilities.

[0028] However, even when predicting trajectories using autoregression methods, further challenges remain. Specifically, as shown in Figure 4, sudden disturbances can occur, causing the actual trajectory to deviate significantly from the prediction. In such cases, as shown in the sudden disturbance model in Figure 4, it is necessary to quickly restore the predicted trajectory to the actual trajectory. Here, the time required from the occurrence of a disturbance until the predicted trajectory returns to the actual trajectory is defined as the "AR recovery time." It is preferable for this AR recovery time to be as short as possible.

[0029] Figure 5 is a graph showing the relationship between the number of AR terms, the error rate, and the AR recovery time. Figure 5 shows the results obtained by the inventors through simulation, assuming a resonant frequency of 80 Hz for the control equipment. As can be seen from Figure 5, the more AR terms there are, the lower the error rate and the better the prediction accuracy. In other words, more actual trajectory data is needed to make highly accurate predictions. On the other hand, the more AR terms there are, the longer the AR recovery time. That is, for sudden disturbances, a larger number of AR terms means that it takes more time to respond, making it unsuitable.

[0030] Therefore, in this disclosure, in a method for predicting orbits using an autoregressive method, the orbit calculation formula used in the autoregressive method is modified when a sudden disturbance occurs and the actual orbit deviates significantly from the prediction. Specifically, when a sudden disturbance occurs, the number of terms in the orbit calculation formula used for orbit prediction is temporarily reduced. This shortens the time it takes for the predicted orbit to return to the actual orbit when a sudden disturbance occurs, enabling a quicker recovery. After the recovery, the prediction accuracy is improved by, for example, returning the orbit calculation formula to its original form or gradually increasing the number of terms in the orbit calculation formula.

[0031] Figure 6 is a block diagram showing an example configuration of a control system according to an embodiment. In the configuration of Figure 6, the trajectory prediction calculation unit 21 calculates the predicted trajectory of the control target 22, for example, the head 11 in a semiconductor chip mounting machine, according to a trajectory calculation formula based on the autoregression method. The trajectory prediction calculation unit 21 is configured to be able to perform both a high-precision mode and a disturbance response mode. In high-precision mode, the trajectory can be predicted with high accuracy by calculating the predicted trajectory according to a trajectory calculation formula with 20 AR terms. On the other hand, in disturbance mode, the recovery time when a disturbance occurs can be shortened by calculating the predicted trajectory according to a trajectory calculation formula with 3 AR terms. Note that the number of AR terms here is merely an example. The control unit 23 controls the movement of the control target 22 along the predicted trajectory calculated by the trajectory prediction calculation unit 21. The predicted trajectory data calculated by the trajectory prediction calculation unit 21 is stored in a prediction data storage unit 26, which is configured, for example, as memory.

[0032] The camera image processing unit 24, which corresponds to the detection unit, for example in a semiconductor chip mounting machine, photographs alignment marks attached to the substrate 14 and detects the current position of the controlled object 22, such as the head 11, from the captured image through image processing. The disturbance determination unit 25 compares the detected position of the controlled object 22 with the prediction data stored in the prediction data storage unit 26 to determine whether or not a disturbance has occurred. For example, if the error between the predicted trajectory and the detected position does not exceed a threshold, it is determined to be normal, and if it exceeds the threshold, it is determined to be abnormal, i.e., a disturbance has occurred. The trajectory prediction calculation unit 21 performs trajectory prediction in high-precision mode when no disturbance has occurred, and performs trajectory prediction in disturbance response mode when a disturbance has occurred.

[0033] The orbital prediction calculation unit 21 and the disturbance detection unit 25 are implemented, for example, by a combination of a program-executable processor and a program for realizing the functions described above.

[0034] Figure 7 is a graph showing the simulation results. The horizontal axis represents the error rate, and the vertical axis represents the recovery time. In Figure 7, "Number of terms 20 / 3" indicates the result of a configuration where the number of AR terms is switched between 20 and 3 depending on the presence or absence of a disturbance, as shown in Figure 6. As a comparative example, Figure 7 shows the results when the number of terms is fixed at 20 and when the number of terms is fixed at 3. As can be seen from Figure 7, with the configuration in Figure 6, the recovery time is significantly shorter than when the number of terms is fixed at 20, and is almost the same as when the number of terms is fixed at 3. Also, the error rate is almost the same as when the number of terms is fixed at 20, and significantly lower than when the number of terms is fixed at 3. In other words, with the configuration in Figure 6, recovery in the event of a disturbance is fast, and the prediction accuracy during normal conditions is high.

[0035] Thus, according to this embodiment, it is possible to perform trajectory predictions that can quickly recover from sudden disturbances while maintaining high prediction accuracy.

[0036] Figure 8 shows an example of a control flow according to an embodiment. Figure 8 assumes a case where control is performed to move the head 11, which is an example of a control target, in a semiconductor chip mounting machine shown in Figure 1.

[0037] First, the position of the head 11 is detected (S11). For example, a camera 15 attached to the head 11 captures an image including alignment marks on the substrate 14. By processing this captured image, the position of the alignment marks is detected, and based on this detection result, the position of the head 11 relative to the substrate 14 is detected.

[0038] The predicted trajectory of head 11 is calculated according to a trajectory calculation formula based on autoregression (S12). Here, for example, a trajectory calculation formula with 10 terms is used. Control is performed to move the position of head 11 according to the calculated predicted trajectory (S13). The predicted trajectory data is also stored in the memory unit (S14).

[0039] After controlling the movement, the position of the head 11 is detected (S15). Then, the predicted trajectory and the detected position of the head 11 are compared, and if the error does not exceed a threshold, it is determined to be normal; if it exceeds a threshold, it is determined to be abnormal (S16).

[0040] If it is determined that the process is normal in step S16, the predicted trajectory of the head is calculated according to the trajectory calculation formula used in step S12 (S17), and control is performed to move the position of the head according to the calculated predicted trajectory (S18). The predicted trajectory data is also stored in the memory unit (S14). Then, the process returns to step S15.

[0041] If an abnormality is detected in step S16, the predicted trajectory of the head is calculated using a trajectory calculation formula with fewer terms than the one used in step S12 (S19). Here, for example, a trajectory calculation formula with 3 terms is used. Control is performed to move the position of the head according to the calculated predicted trajectory (S1A). The predicted trajectory data is also stored in the memory unit (S14). Then, the process returns to step S15.

[0042] The control flow shown in Figure 8 allows for high accuracy in predicting the trajectory of the head 11 while enabling rapid recovery from sudden disturbances. Therefore, high-precision and stable control can be achieved in semiconductor chip mounting machines.

[0043] In the flowchart in Figure 8, when an abnormality is determined in step S16, a trajectory calculation formula with 3 terms is used. However, the trajectory calculation formula used when an abnormality is determined is not limited to this. In other words, when an abnormality is determined, a formula with fewer terms than the one used during normal operation should be used.

[0044] Furthermore, multiple trajectory calculation formulas may be prepared before controlling the movement of the head 11. In this case, for example, when an abnormality is detected in step S16, the trajectory calculation formula that minimizes the error between the predicted trajectory and the detected position of the head 11 may be selected from among the prepared multiple trajectory calculation formulas, and the trajectory prediction may be performed using the selected trajectory calculation formula. This enables a faster recovery.

[0045] Furthermore, in the flow chart of Figure 8, if an abnormality is determined in step S16 and the trajectory prediction is performed using a trajectory calculation formula with 3 terms, and then it is determined to be normal, the original trajectory calculation formula with 10 terms is used. However, the control flow is not limited to this, and for example, the number of terms in the trajectory calculation formula may be gradually increased. Specifically, for example, if an abnormality is determined and the trajectory prediction is performed using a trajectory calculation formula with 3 terms, and then it is determined to be normal, a trajectory calculation formula with more terms than 3 and fewer terms than the original trajectory calculation formula with 10 terms, for example, a trajectory calculation formula with 5 terms, may be used. Subsequently, if it is determined to be normal, a trajectory calculation formula with 7 terms may be used, and if it is determined to be normal again, the original trajectory calculation formula with 10 terms may be used, and so on.

[0046] Figure 9 shows an example of control after an abnormality is detected. In the flow of Figure 9, when an abnormality is detected, a trajectory prediction is performed using a trajectory calculation formula with fewer terms (S21), and control is performed to move the position of the head according to the calculated predicted trajectory (S22). After that, the position of the head 11 is detected (S23).

[0047] Then, trajectory prediction is performed according to the same trajectory calculation formula as in step S21 (S24), and prediction data B is obtained. Also, trajectory prediction is performed according to a trajectory calculation formula with one more term than the trajectory calculation formula used in step S21 (S25), and prediction data C is obtained. The error between the actual trajectory detected in step S23 and prediction data B and prediction data C is calculated and compared (S26). If the error of prediction data B is smaller, control is performed to move the head position according to prediction data B (S27). On the other hand, if the error of prediction data C is smaller, control is performed to move the head position according to prediction data C (S28). Then, the process returns to step S23.

[0048] With the control shown in Figure 9, if the error between the predicted trajectory and the detected position is large and an anomaly is detected, the number of terms in the trajectory calculation formula is reduced, and then gradually increased. This allows the predicted trajectory to be restored to the actual trajectory more stably. In the control flow of Figure 9, the number of terms in the trajectory calculation formula is increased by one each time, but this is not the only option; for example, the number of terms could be increased by two each time.

[0049] Furthermore, in the trajectory prediction method according to this embodiment, the user may set the number of terms and coefficients in the trajectory calculation formula, the threshold for determining the occurrence of disturbances, etc., via a user interface before the movement of the controlled object.

[0050] Figure 10 shows an example of a user interface screen for the control system. Before moving the head 11, the user may set the number of terms A1 and coefficient A2 in the trajectory calculation formula from a user interface such as the one shown in Figure 10. The user may also set a threshold A3 used to determine the occurrence of disturbances.

[0051] Furthermore, when the control system determines that a disturbance has occurred, it may output a message indicating that a disturbance has occurred, along with the trajectory calculation formula used to calculate the predicted trajectory, via a user interface as shown in Figure 10.

[0052] Furthermore, in this embodiment, it is preferable to use model predictive control to control the movement of the head 11. Figure 11 is a diagram illustrating model predictive control. As shown in Figure 11, model predictive control commands control that anticipates the expected operation by the control. The control command is obtained by multiplying the error between the target and the current by the inverse matrix of the optimized response coefficient.

[0053] The functions realized by the components described herein may be implemented in a circuitry or processing circuitry, including general-purpose processors, application-specific processors, integrated circuits, ASICs (Application Specific Integrated Circuits), CPUs (a Central Processing Unit), conventional circuits, and / or combinations thereof, programmed to realize the functions described herein. A processor includes transistors and other circuits and is considered a circuitry or processing circuitry. A processor may be a programmed processor that executes a program stored in memory.

[0054] In this specification, circuitry, unit, and means are hardware programmed to perform or execute the functions described herein. Such hardware may be any hardware disclosed herein, or any hardware known to be programmed to perform or execute the functions described herein.

[0055] If the hardware is a processor that is considered to be a type of circuitry, then the circuitry, means, or unit is a combination of hardware and software used to constitute the hardware and / or processor. [Industrial applicability]

[0056] This disclosure enables rapid recovery from sudden disturbances while maintaining high prediction accuracy for the trajectory prediction of a controlled object, and is therefore useful, for example, for improving the operational accuracy of semiconductor chip mounting machines. [Explanation of Symbols]

[0057] 11. Head (Controlled object) 21 Orbital Prediction Calculation Unit 22 Controlled object 23 Control Unit 24 Camera image processing unit (detection unit) 25 Disturbance detection unit

Claims

1. A method for predicting the trajectory of a controlled object using an autoregressive method, and moving the controlled object according to the predicted trajectory, Step (a) calculates the predicted trajectory of the controlled object according to the first trajectory calculation formula and moves the controlled object along the calculated predicted trajectory, (b) a step of detecting the position of the moved control object, Step (c) involves comparing the predicted trajectory calculated in step (a) with the position of the controlled object detected in step (b), determining that it is normal if the error between the predicted trajectory and the detected position does not exceed a threshold, and determining that it is abnormal if it exceeds the threshold. If it is determined to be normal in step (c), step (d) moves the controlled object along the predicted trajectory according to the first trajectory calculation formula, If an abnormality is determined in step (c), the system includes step (e) of calculating a predicted trajectory according to a second trajectory calculation formula which has fewer terms than the first trajectory calculation formula, and moving the controlled object along the calculated predicted trajectory. Orbit prediction method.

2. In the trajectory prediction method according to claim 1, After step (e), steps (b) and (c) are performed again. If the system determines that step (c) is normal when executed again, it calculates the predicted trajectory according to the first trajectory calculation formula or a third trajectory calculation formula which has more terms than the second trajectory calculation formula, and moves the controlled object along the calculated predicted trajectory. Orbit prediction method.

3. In the trajectory prediction method according to claim 1, The process includes a step (f) of preparing multiple trajectory calculation formulas before the movement of the controlled object, In step (e), from among the multiple trajectory calculation formulas prepared in step (f), the trajectory calculation formula that minimizes the error between the predicted trajectory and the detected position of the controlled object is selected as the second trajectory calculation formula. Orbit prediction method.

4. In the trajectory prediction method according to claim 1, Prior to the movement of the controlled object, the user sets the number of terms and coefficients in the trajectory calculation formula, as well as the threshold, via the user interface. Orbit prediction method.

5. In the trajectory prediction method according to claim 1, Step (b) is performed based on the image captured by the imaging device. Orbit prediction method.

6. A control system that predicts the trajectory of a controlled object using an autoregressive method and moves the controlled object according to the predicted trajectory, A detection unit for detecting the position of the controlled object, A calculation unit that calculates the predicted trajectory of the controlled object according to the trajectory calculation formula, A control unit controls the movement of the controlled object along the predicted trajectory calculated by the calculation unit, The system includes a disturbance determination unit that compares the predicted trajectory calculated by the calculation unit with the detected position detected by the detection unit, and determines that a disturbance has occurred if the error between the predicted trajectory and the detected position exceeds a threshold. The calculation unit calculates the predicted trajectory according to the first trajectory calculation formula, and when the disturbance determination unit determines that a disturbance has occurred, it calculates the predicted trajectory according to the second trajectory calculation formula, which has fewer terms than the first trajectory calculation formula. Control system.

7. In the control system according to claim 6, The system includes a user interface that allows the user to set the number of terms and coefficients in the orbital calculation formula, as well as the threshold value. When the disturbance detection unit determines that a disturbance has occurred, the control system outputs, via the user interface, a statement indicating that a disturbance has occurred and the orbital calculation formula used to calculate the predicted orbit. Control system.

Images