Automatic train operation system and automatic train operation method

Abstract

To obtain an automatic train operation system capable of achieving both ride comfort and control responsiveness while suppressing influence caused by detection errors of a running state.SOLUTION: An automatic train operation system 10 includes a running state calculation unit 11 that calculates a running state of a train 20, a running plan information holding unit 12 that holds running plan information of the train 20, a dynamic characteristic model holding unit 13 that holds a dynamic characteristic model of a driving device 25, and a command value generation unit 14 that calculates a commanded acceleration for following the running plan information and converts the commanded acceleration into a notch command to output the notch command. The command value generation unit 14 extracts a target trajectory from the running plan information, sets constraint conditions, calculates a predicted running state within a range of the constraint conditions on the basis of the dynamic characteristic model, calculates a commanded acceleration change amount that minimizes, within the range of the constraint conditions, an evaluation value of a sum of a square-sum of a difference between the predicted running state and the target trajectory and an absolute value sum of the commanded acceleration change amounts, and calculates the commanded acceleration using the commanded acceleration change amount.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to an automatic train operation system and an automatic train operation method for controlling train operation. [Background technology]

[0002] Conventionally, automatic train operation systems have the function of setting a target run curve (hereinafter referred to as the target trajectory), calculating an optimal command value (specifically, powering acceleration or braking deceleration) to track the target trajectory, converting it into a notch command, and outputting it to control speed. A known method of calculating the command value is to calculate a future running state (hereinafter referred to as the predicted running state) based on the current detected running state (e.g., speed, acceleration, position, etc.) and the command value, setting an evaluation function using the "deviation between the predicted running state and the target trajectory" and the "command value change amount," and then calculating a command value that minimizes the set evaluation function, thereby achieving a command value that minimizes acceleration change, i.e., a comfortable ride, and enables tracking of the target trajectory. For example, Patent Document 1 discloses a technology that sets an evaluation function consisting of a "sum of squares of the deviation between the predicted running state and the target trajectory" and a "sum of squares of the command value change amount," and then calculates a command value. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2000-156920 Summary of the Invention [Problem to be solved by the invention]

[0004] However, with the above-described conventional technology, when the evaluation value of the "sum of squares of the deviation between the predicted driving state and the target trajectory" fluctuates due to a detection error of the driving state, the command value calculated by the calculation varies significantly. When converting a command value to a notch command, if the command value fluctuates, particularly near the conversion threshold, hunting occurs in the notch command, causing the notch to switch frequently in a short period of time, resulting in a poor ride. To prevent hunting due to the above-described detection error, the above-described conventional technology proposes providing a dead zone in the conversion characteristic from the command value to the notch command, so that the notch is maintained for a while after switching. However, providing such a dead zone results in a decrease in responsiveness, such as a delay in switching the notch command in situations where the command value must be changed immediately, such as when braking to avoid exceeding the speed limit.

[0005] The present disclosure has been made in consideration of the above, and aims to provide an automatic train operation system that can achieve both ride comfort and control responsiveness while suppressing the effects of detection errors in running conditions. [Means for solving the problem]

[0006] To solve the above-mentioned problems and achieve the object, the present disclosure provides an automatic train operation system that controls the speed of a train by using notch commands to a driving device that applies driving force to the train. The automatic train operation system includes a running state calculation unit that calculates the current running state of the train, a trip plan information storage unit that stores trip plan information for the train between stations, a dynamic characteristics model storage unit that stores a dynamic characteristics model of the driving device, and a command value generation unit that calculates a command acceleration to follow the trip plan information, converts the command acceleration into a notch command, and outputs the notch command. The command value generation unit extracts a target trajectory from the trip plan information up to a time ahead specified from a time step of the current position of the train, sets constraints on predicted running states and command acceleration change amounts when the command acceleration is changed, calculates predicted running states within the constraint conditions based on the dynamic characteristics model, uses an evaluation function that is the sum of the sum of the squares of the differences between the predicted running states and the target trajectory and the sum of the absolute values ​​of the command acceleration change amounts, calculates a command acceleration change amount that minimizes the evaluation value of the evaluation function within the constraint conditions, and calculates the command acceleration using the command acceleration change amount. [Effects of the Invention]

[0007] According to the present disclosure, an automatic train operation system has the advantage of being able to achieve both ride comfort and control responsiveness while suppressing the effects of detection errors in the running state. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is a diagram showing a configuration example of an automatic train operation system according to a first embodiment. [Figure 2] FIG. 1 is a diagram showing an example of operation plan information stored in an operation plan information storage unit included in the automatic train operation system according to the first embodiment. [Figure 3] FIG. 1 is a diagram showing an image of a dynamic characteristics model stored in a dynamic characteristics model storage unit included in the automatic train operation system according to the first embodiment. [Figure 4] FIG. 1 is a diagram showing a configuration example of a command value generation unit included in the automatic train operation system according to the first embodiment. [Figure 5]FIG. 10 is a diagram illustrating an image of the operation of a target trajectory extraction unit of a command value generation unit included in the automatic train operation system according to the first embodiment. [Figure 6] FIG. 10 is a diagram showing a method for calculating a predicted acceleration by a command acceleration change amount calculation unit of a command value generation unit included in the automatic train operation system according to the first embodiment. [Figure 7] FIG. 10 is a diagram showing an operation of a notch command conversion unit of a command value generation unit included in the automatic train operation system according to the first embodiment, converting a command acceleration into a notch command. [Figure 8] FIG. 1 shows the characteristics of the evaluation function set in Patent Document 1 as a comparative example. [Figure 9] FIG. 10 is a diagram showing, as a comparative example, fluctuations in notch commands when command values ​​vary greatly. [Figure 10] FIG. 1 is a diagram showing characteristics of an evaluation function set in a command value generation unit included in the automatic train operation system according to the first embodiment. [Figure 11] FIG. 10 is a diagram showing an example of a target trajectory and a predicted running state when a command value generating unit provided in an automatic train operation system assigns a certain weighting as a comparative example. [Figure 12] FIG. 10 is a diagram showing an example of a target trajectory and a predicted running state when a command value generating unit included in the automatic train operation system according to the first embodiment changes the magnitude of weighting according to the time from the present. [Figure 13] Flowchart showing the operation of the automatic train operation system according to the first embodiment [Figure 14] A flowchart showing the operation of a command value generation unit included in the automatic train operation system according to the first embodiment. [Figure 15] FIG. 1 is a diagram showing an example in which a processing circuit for realizing the automatic train operation system according to the first embodiment is configured with a processor and a memory. [Figure 16] FIG. 1 is a diagram showing an example in which a processing circuit for realizing the automatic train operation system according to the first embodiment is configured with dedicated hardware. [Figure 17] FIG. 10 is a diagram showing a configuration example of a command value generation unit included in an automatic train operation system according to a second embodiment. [Figure 18]A flowchart showing the operation of a command value generating unit included in an automatic train operation system according to a second embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0009] An automatic train operation system and an automatic train operation method according to an embodiment of the present disclosure will be described in detail below with reference to the accompanying drawings.

[0010] Embodiment 1 FIG. 1 is a diagram showing an example of the configuration of an automatic train operation system 10 according to the first embodiment. The automatic train operation system 10 is mounted on a train 20 and controls the operation of the train 20. The automatic train operation system 10 controls the speed of the train 20 by issuing a notch command to a driving device 25 that provides driving force for accelerating or decelerating the train 20. The automatic train operation system 10 includes a running state calculation unit 11, a running plan information storage unit 12, a dynamic characteristics model storage unit 13, and a command value generation unit 14.

[0011] The running state calculation unit 11 calculates the current running state of the train 20. Specifically, the running state calculation unit 11 acquires a speed signal indicating a value corresponding to the current speed of the train 20 from a tachometer generator 21 mounted on the train 20. The running state calculation unit 11 also acquires, from an on-board terminal 22 mounted on the train 20, point information, which is information about the installation point of the beside terminal 31 acquired by the on-board terminal 22 from the beside terminal 31. The point information may be information indicating the installation point of the beside terminal 31 itself, or may be identification information of the beside terminal 31 in a case where the installation position of the beside terminal 31 can be identified by referring to information indicating the relationship between the identification information of the beside terminal 31 and the installation position of the beside terminal 31. The fact that the running state calculation unit 11 has acquired point information from the on-board terminal 22 means that the on-board terminal 22 mounted on the train 20 has passed the point where the beside terminal 31 is installed at that time. The running state calculation unit 11 calculates the current running state x using the speed signal acquired from the speed generator 21 and the location information acquired from the on-board terminal 22, where k is the time step after departure from the station. k Calculate the driving state x k is expressed as the position pk , velocity v k , acceleration a k Includes:

[0012]

number

[0013] From now on, running state x k At position p k , velocity v k , and acceleration a k However, other information such as power consumption may also be included. k has an acceleration a k The running state calculation unit 11 calculates a running state including at least the current position and the current speed of the train 20, using the speed signal acquired from the tachometer generator 21 and the location information of the train 20 obtained from information acquired by the on-board coil 22 from the ground coil 31. The running state calculation unit 11 outputs the calculated running state to the command value generation unit 14.

[0014] The trip plan information storage unit 12 stores trip plan information for the train 20 between stations where the train 20 is running. The trip plan information storage unit 12 stores at least the running time, target position, target speed, and speed limit for the train 20 between stations where the train 20 is running as trip plan information. The trip plan information storage unit 12 is, for example, a database stored in an internal data memory of the automatic train operation system 10, and stores trip plan information for each station where the train 20 is running. The trip plan information for each station where the train 20 is running is, for example, as shown in FIG. 2, a target position p(^) corresponding to a time step k of a running time k. k , target speed v(^) k , reference speed limit v(￣) k , target acceleration a(^) k , etc.

[0015] FIG. 2 is a diagram showing an example of operation plan information stored in the operation plan information storage unit 12 included in the automatic train operation system 10 according to the first embodiment. The table on the left side of FIG. 2 shows operation plan information for each station between which the train 20 runs. The graph on the right side of FIG. 2 shows an image of the transition of the target position, target speed, and target acceleration shown in the table on the left side. Reference speed limit v( ) k is the speed limit that is constantly set at each point between stations depending on track conditions, vehicle performance, etc., and is assumed to be the ATC (Automatic Train Control) speed limit, for example. Track conditions include curves, downward gradients, switches, and the type of track section. Reference speed limit v(￣) k The travel plan information may include linear information such as gradient and curvature.

[0016] Here, the specification cannot express the state where p has a ^ above it, so the specification expresses the state where p has a ^ above it as p(^). Similarly, the specification cannot express the state where v has a ￣ above it, so the specification expresses the state where v has a ￣ above it as v(￣).

[0017] The trip plan information is stored in advance in an internal data memory of the automatic train operation system 10 before the train 20 starts operating. The automatic train operation system 10 may be provided with a separate function for creating a target trip curve, similar to the train trip control device described in JP 2004-229359 A and the like, and may create a trip plan between stations while the train 20 is stopped at a station, or create a trip plan while the train 20 is traveling, and store the trip plan information in the trip plan information storage unit 12. The trip plan information storage unit 12 may store the trip plan information in a function format with respect to the traveling time. By storing the trip plan information in a function format, the trip plan information storage unit 12 can store the trip plan information with a small amount of data.

[0018] The dynamic characteristics model holding unit 13 holds a dynamic characteristics model of the drive unit 25 mounted on the train 20. The dynamic characteristics model holding unit 13 is, for example, a database stored in an internal data memory of the automatic train operation system 10, and holds the dynamic characteristics model of the drive unit 25. The dynamic characteristics model is the response characteristics of the drive acceleration of the drive unit 25 to a command acceleration, which is a command value generated by the automatic train operation system 10, and represents the torque characteristics of the motor mounted on the train 20 and the braking characteristics of the air brake, regenerative brake, etc., using model parameters. FIG. 3 is a diagram showing an example of the dynamic characteristics model held by the dynamic characteristics model holding unit 13 included in the automatic train operation system 10 according to the first embodiment. The model parameters include, for example, a response delay of the drive acceleration when the command acceleration changes, the rise speed and fall speed of the drive acceleration, etc. The dynamic characteristics model shown in FIG. 3 is an example, and the dynamic characteristics model holding unit 13 may hold other model parameters as the dynamic characteristics model.

[0019] In this embodiment, as shown in FIG. 1 , it is assumed that the automatic train operation system 10 outputs a notch command directly to the drive unit 25. However, there may be other devices, such as a monitor device (not shown), between the automatic train operation system 10 and the drive unit 25. Therefore, if there are other devices between the automatic train operation system 10 and the drive unit 25, the dynamic characteristics model holding unit 13 may hold, as a dynamic characteristics model, response delays and other delays resulting from transmission processing when the notch command passes through the other devices. Furthermore, the running conditions of the train 20 may also affect the response delay when the command acceleration changes, the rise speed and fall speed of the drive acceleration, and other factors, such as when the train 20 accelerates while traveling uphill, when the train 20 decelerates while traveling downhill, when the train 20 is traveling in a curved section, or when the train 20 is significantly affected by air resistance during travel. Therefore, the dynamic characteristics model holding unit 13 may hold a dynamic characteristics model according to the running conditions of the train 20.

[0020] The command value generation unit 14 calculates a command acceleration, which is a command value for following the trip plan information, converts the command acceleration into a notch command, and outputs it to the drive device 25. The detailed configuration and operation of the command value generation unit 14 will be described. FIG. 4 is a diagram showing an example configuration of the command value generation unit 14 provided in the automatic train operation system 10 according to the first embodiment. The command value generation unit 14 includes a target trajectory extraction unit 141, a state constraint generation unit 142, a command acceleration change amount calculation unit 143, a command acceleration calculation unit 144, and a notch command conversion unit 145.

[0021] When the target trajectory extraction unit 141 acquires the running state from the running state calculation unit 11, it extracts a target trajectory from the running plan information stored in the running plan information storage unit 12 up to a time ahead specified from the time step of the current position of the train 20.

[0022] FIG. 5 is a diagram illustrating an image of the operation of the target trajectory extraction unit 141 of the command value generation unit 14 provided in the automatic train operation system 10 according to the first embodiment. The upper table in FIG. 5 illustrates an image of the operation of the target trajectory extraction unit 141 extracting information up to a required time ahead as a target trajectory from the operation plan information stored in the operation plan information storage unit 12. The lower graph in FIG. 5 illustrates an image of the transition of the target position, target acceleration, target speed, and reference speed limit extracted as a target trajectory from the upper table. The target trajectory extraction unit 141 refers to operation plan information between the current operation stations, and extracts, as a target trajectory, each state quantity of the operation plan information for L time steps ahead from the time step of the current position, i.e., from time step n to time step n+L in the example of FIG. 5. Note that L is an integer equal to or greater than 0. The target trajectory (r n ,…,r n+i ,…,r n+L ) is the target position (p(^) n ,…,p(^) n+i ,…,p(^) n+L ), target speed (v(^) n ,…,v(^) n+i ,…,v(^) n+L ), and the target acceleration (a(^) n ,…,a(^) n+i ,…,a(^) n+L) and the target trajectory (r n ,…,r n+i ,…,r n+L ) is expressed as in equation (2).

[0023]

number

[0024] Furthermore, the target trajectory extraction unit 141 similarly refers to the running plan information between the current running stations, and calculates the reference speed limit (v(￣)) for the L time steps ahead from the time step of the current position. n ,…,v(￣) n+i ,…,v(￣) n+L The target trajectory extraction unit 141 outputs the target trajectory to the command acceleration change amount calculation unit 143, and outputs the reference limit speed to the state constraint generation unit 142. Note that the time width of each time step is assumed to be, for example, about several hundred ms, but is not limited to this.

[0025] The state constraint generation unit 142 acquires the reference speed limit from the target trajectory extraction unit 141 and acquires the speed limit of the train 20 from the ATC on-board equipment 24. In the train 20, as shown in FIG. 1 , the power receiver 23 receives the speed limit transmitted from the ATC ground equipment 32 via the rail 33. The ATC on-board equipment 24 outputs the speed limit received by the power receiver 23 to the state constraint generation unit 142. The state constraint generation unit 142 sets constraint conditions on the predicted running state and commanded acceleration change amount used in the calculation by the commanded acceleration change amount calculation unit 143, using the reference speed limit acquired from the target trajectory extraction unit 141 and the speed limit acquired from the ATC on-board equipment 24. The state constraint generation unit 142 sets the following three constraint conditions on the predicted running state and commanded acceleration change amount when the commanded acceleration change amount calculation unit 143 minimizes the evaluation value V of an evaluation function described later.

[0026] First, the state constraint generator 142 calculates the predicted speed v k+i is the reference speed limit v(￣) k+j The constraint condition shown in equation (3) is set so that the predicted speed vk+i is the speed limit v(￣) based on the signal aspect of the ATC on-board device 24 ATC The constraints shown in equation (4) are set so that the following holds: The speed limit is set to v(￣) based on the signal aspect with a margin. ATC By adding equation (4) to equation (3), which is the constraint condition related to the steady-state speed limit, it is possible to deal with sudden speed limits, such as a drop in the signal level due to a block section depending on the location of a preceding train (not shown).

[0027]

number

number

[0028] Second, the state constraint generator 142 calculates the predicted acceleration a k+j is within a specified range, e.g., the minimum value a Lo to maximum value a Up The constraint condition shown in equation (5) is set so that the predicted acceleration a k+j The upper and lower limits are set to the performance limits of the output acceleration / deceleration of the drive unit 25, but may be set to a range narrower than the performance limits in consideration of the ride comfort.

[0029]

number

[0030] Third, the state constraint generator 142 calculates the command acceleration change amount ΔU k+j is within a specified range, e.g., the minimum value ΔU Lo to maximum value ΔU Up The constraint condition shown in equation (6) is set so that the command acceleration change amount ΔU is within the range of k+j The upper and lower limits are set to the performance limits of the amount of change in the output acceleration / deceleration of the drive unit 25, but may be set to a range narrower than the performance limits in consideration of the ride comfort.

[0031]

number

[0032] The state constraint generating unit 142 outputs the set constraint conditions to the command acceleration change amount calculating unit 143 .

[0033] The command acceleration change amount calculation unit 143 calculates a predicted running state when the command acceleration is changed based on the dynamic characteristics model acquired from the dynamic characteristics model holding unit 13 within the range of the constraint conditions, and calculates a command acceleration change amount that minimizes the evaluation value V of the above-mentioned evaluation function within the range of the constraint conditions. Specifically, the command acceleration change amount calculation unit 143 calculates a command acceleration change amount (ΔU k ,…,U k+j ,…,ΔU k+M ) is set, and the predicted driving state (x k ,…,x k+i ,…,x k+L ) is calculated. M is an integer equal to or greater than 0, and M≦L. k ,…,x k+i ,…,x k+L ) is the current driving state x k and the predicted position (p k+1 ,…,p k+i ,…,p k+L ), predicted speed (v k+1 ,…,v k+i ,…,v k+L ), and predicted acceleration (a k+1 ,…,a k+i ,…,a k+L ) and the predicted driving state (x k ,…,x k+i ,…,x k+L ) is expressed as in equation (7).

[0034]

number

[0035] Predicted acceleration a k+iis the actual command acceleration U at the previous time step k-1 k-1 , command acceleration change amount ΔU k+j , and the current acceleration a k 6 is a graph showing the predicted acceleration a calculated by the command acceleration change amount calculation unit 143 of the command value generation unit 14 included in the automatic train operation system 10 according to the first embodiment. k+i The command acceleration change amount calculation unit 143 calculates the actual command acceleration U k-1 The command acceleration change amount calculation unit 143 obtains the predicted speed v from the notch command conversion unit 145 as described below. k+i At each time step, the predicted acceleration is integrated over time to obtain the current velocity v k The command acceleration change amount calculation unit 143 calculates the predicted position p k+i Similarly, for each time step, the predicted velocity is integrated over time to obtain the current position p k It should be noted that in FIG. 6, the change in the prediction line based on the model parameters, which indicates the transition of acceleration, is not constant, but this is because the response delay and the like explained in FIG. 3 are taken into consideration. However, the prediction line based on the model parameters shown in FIG. 6 is only an example, and is not limited to this shape.

[0036] The command acceleration change amount calculation unit 143 calculates the predicted running state x at each time step as shown in equation (8). k+i and the target trajectory r n+i The weighting coefficient Q for each driving state and time step for the deviation from i The sum of squares of the deviation between the predicted running state and the target trajectory, which is the sum of squares multiplied by , and the command acceleration change amount ΔU k+j The absolute value of the weighting factor λ j The sum of the absolute value of the command acceleration change amount is set as an evaluation function of the sum of the absolute value of the command acceleration change amount multiplied by ...

[0037]

number

[0038] In equation (8), ( ) T indicates the transpose of the matrix in parentheses. i The larger the time step, the smaller the weighting coefficient Q i is expressed as shown in equation (9).

[0039]

number

[0040] In equation (9), diag() indicates that the elements in the parentheses are diagonal matrices. Also, in equation (9), Q p,i is the weighting factor for the difference between the predicted position and the target position at each time step, and Q v,i is the weighting factor for the difference between the predicted speed and the target speed at each time step, and Q a,i is the weighting coefficient for each time step for the difference between the predicted acceleration and the target acceleration. k+j Weighting coefficient λ for the absolute value of j Regarding the weighting factor Q i Similarly, the weighting value may be changed or may be a constant value.

[0041] The command acceleration change amount calculation unit 143 calculates the command acceleration change amount (ΔU k ,…,ΔU k+J ,…,ΔU k+M The command acceleration change amount calculation unit 143 calculates the command acceleration change amount (ΔU k ,…,ΔU k+J ,…,ΔU k+M ) to the command acceleration calculation unit 144.

[0042] The command acceleration calculation unit 144 receives the command acceleration change amount (ΔU k ,…,ΔU k+J ,…,ΔU k+M ) and obtains the actual command acceleration from the notch command conversion unit 145. The command acceleration calculation unit 144 calculates the command acceleration change amount (ΔUk ,…,ΔU k+j ,…,ΔU k+M Specifically, the command acceleration calculation unit 144 calculates the command acceleration change amount ΔU as shown in equation (10). k+j and the actual command acceleration U at the previous time step k-1 k-1 and calculate the command acceleration plan (U k ,…,U k+M ) is calculated.

[0043]

number

[0044] The command acceleration calculation unit 144 calculates the command acceleration plan (U k ,…,U k+M ) is the value of the command acceleration U k is output to the notch command conversion unit 145 as the command acceleration at the current time.

[0045] The notch command conversion unit 145 acquires information on the speed of the train 20 included in the running state from the running state calculation unit 11, and calculates the command acceleration U k The notch command conversion unit 145 obtains the command acceleration U k is converted into a notch command and output to the driving device 25. Specifically, as shown in FIG. 7, the notch command conversion unit 145 converts the command acceleration U k 7 shows how the notch command conversion unit 145 of the command value generation unit 14 included in the automatic train operation system 10 according to the first embodiment converts the command acceleration U k 7 is a diagram showing an operation of converting the acquired speed and command acceleration U k By selecting a notch command that matches kThe notch command conversion unit 145 converts the acceleration value corresponding to the notch command output to the driving device 25 into an actual command acceleration, and outputs the actual command acceleration to the command acceleration change amount calculation unit 143 and the command acceleration calculation unit 144.

[0046] In this way, in the automatic train operation system 10, the command value generation unit 14 extracts a target trajectory from the running plan information up to a specified time ahead from the time step of the current position of the train 20, sets constraints on the predicted running state and the command acceleration change amount when the command acceleration is changed, calculates the predicted running state within the range of the constraints based on the dynamic characteristics model, uses the sum of the squared difference between the predicted running state and the target trajectory and the sum of the absolute values ​​of the command acceleration change amount as an evaluation function, calculates the command acceleration change amount that minimizes the evaluation value V of the evaluation function within the range of the constraints, and calculates the command acceleration using the command acceleration change amount.

[0047] In this embodiment, command value generating unit 14 sets an evaluation function consisting of a "sum of squares of deviation between predicted traveling state and target trajectory" and a "sum of absolute values ​​of command acceleration change amount," and calculates a command acceleration change amount that minimizes evaluation value V of the evaluation function. The minimization process of evaluation value V, which is an evaluation function using the "sum of absolute values ​​of command acceleration change amount," has the effect of reducing the command value change amount to zero as much as possible, compared to the minimization process of an evaluation function using the "sum of squares of command value change amount" described in Patent Document 1.

[0048] FIG. 8 shows the characteristics of the evaluation function established in Patent Document 1 as a comparative example. In FIG. 8, FIG. 8(a) shows the transition of the evaluation function when the detection error is small, and FIG. 8(b) shows the transition of the evaluation function when the detection error increases. In the comparative example, as the detection error increases, the deviation in the command value change amount increases due to the influence of the deviation in the sum of squares of the deviation between the predicted driving state and the target trajectory, as shown in FIG. 8(b), and the command value obtained by calculation also varies greatly. If the command value varies greatly when converting from the command value to a notch command, especially if the command value fluctuates near the conversion threshold, as shown in FIG. 9, hunting occurs in the notch command, i.e., the notch command frequently switches in a short period of time, resulting in poor ride comfort. FIG. 9 shows the fluctuation of the notch command when the command value fluctuates greatly as a comparative example. As shown in FIG. 9(a), the command value fluctuates near the conversion threshold between notch A and notch B, resulting in frequent switching of the notch command in a short period of time, as shown in FIG. 9(b).

[0049] FIG. 10 is a diagram illustrating the characteristics of the evaluation function set by the command value generator 14 included in the automatic train operation system 10 according to the first embodiment. In FIG. 10, FIGS. 10(a) and 10(b) are based on the same conditions as those in FIGS. 8(a) and 8(b), respectively. However, as shown in FIG. 10(b), even when the detection error increases and the value of the "sum of squares of the deviation between the predicted running state and the target trajectory" fluctuates, the command acceleration change amount does not vary. As a result, this embodiment can prevent hunting of the notch command due to detection errors. This allows the automatic train operation system 10 to prevent hunting without providing a dead zone in the conversion characteristics to the notch command, thereby achieving both ride comfort and responsiveness.

[0050] In this embodiment, the command value generating unit 14 also applies a weighting coefficient Q to the evaluation function that calculates the command acceleration change amount for each time step with respect to the difference between the predicted running state from the present to the Lth time step ahead and the target trajectory. iThe sum of squares multiplied by the sum of squares is used to calculate the command acceleration change amount that minimizes the evaluation value V within the range of the speed limit constraint on the target trajectory. Here, in order to follow the target trajectory by taking into account the speed limit further ahead and reducing the change in command acceleration, it is desirable to set the time step L large, that is, to use many time steps. On the other hand, if the time step L is set large, a running plan for a wider section will be referenced, and for example, as shown in Figure 11, a target trajectory that is greatly curved will be set as the following target. Figure 11 is a diagram showing an example of a target trajectory and a predicted running state when the command value generation unit 14 provided in the automatic train operation system 10 assigns a certain weighting, as a comparative example. In such a case, the weighting coefficient Q i If is set flat, the command acceleration is set to assume the average trajectory in order to track the target trajectory with less change in the command acceleration, resulting in a larger tracking error. In Figure 11, the predicted driving state is unable to track the target trajectory.

[0051] In this embodiment, the command value generator 14 maintains the constraints while adjusting the weighting coefficient Q i The larger the time step, the smaller the weighting coefficient Q i The weighting coefficient Q i By reducing , priority is given to nearby tracking ability and long-distance tracking error is tolerated. FIG. 12 is a diagram showing an example of a target trajectory and a predicted running state when the command value generation unit 14 included in the automatic train operation system 10 according to the first embodiment changes the magnitude of the weighting according to the time from the present. Even if a command acceleration is currently assumed to have a long-distance tracking error, it is expected that the tracking error will be eliminated when the command acceleration is recalculated over time. This allows the command value generation unit 14 to track the target trajectory while looking ahead to ensure that the predicted running state does not exceed the speed limit over a long distance. In FIG. 12, the predicted running state can track the target trajectory.

[0052] In this way, in command value generating unit 14, when command acceleration change amount calculating unit 143 weights the square of the difference between the predicted driving state and the target trajectory at each time step, the weight is increased as the square of the difference between the predicted driving state and the target trajectory at a time step closer to the present. Command acceleration change amount calculating unit 143 also weights the absolute value of the command acceleration change amount at each time step, but the weighting value may be changed depending on the time step or may be a constant value.

[0053] 13 is a flowchart showing the operation of the automatic train operation system 10 according to the embodiment 1. In the automatic train operation system 10, the running state calculation unit 11 calculates the current running state of the train 20 (step S11). The command value generation unit 14 generates a command acceleration for following the running plan information, converts the command acceleration into a notch command, and outputs the notch command (step S12).

[0054] FIG. 14 is a flowchart showing the operation of the command value generation unit 14 included in the automatic train operation system 10 according to the first embodiment. The flowchart shown in FIG. 14 shows in detail the operation of step S12 in the flowchart shown in FIG. 13. In the command value generation unit 14, the target trajectory extraction unit 141 extracts a target trajectory from the trip plan information stored in the trip plan information storage unit 12 to a time ahead specified from the time step of the current position of the train 20 (step S21). The state constraint generation unit 142 sets constraint conditions for the predicted running state and the command acceleration change amount (step S22). The command acceleration change amount calculation unit 143 calculates a predicted running state when the command acceleration is changed based on the dynamic characteristics model, and calculates a command acceleration change amount that minimizes the evaluation value V of the evaluation function within the range of the constraint conditions (step S23). The command acceleration calculation unit 144 calculates the command acceleration using the command acceleration change amount (step S24). The notch command conversion unit 145 converts the command acceleration U k is converted into a notch command and output to the driving device 25 (step S25).

[0055] Next, the hardware configuration of the automatic train operation system 10 will be described. In the automatic train operation system 10, the running plan information storage unit 12 and the dynamic characteristics model storage unit 13 are realized by memory. The running state calculation unit 11 and the command value generation unit 14 are realized by processing circuits. The processing circuits may be a processor and memory that executes a program stored in memory, or may be dedicated hardware.

[0056] FIG. 15 is a diagram illustrating an example in which a processing circuit 90 that realizes the automatic train operation system 10 according to the first embodiment is configured with a processor 91 and a memory 92. When the processing circuit 90 is configured with the processor 91 and the memory 92, each function of the processing circuit 90 of the automatic train operation system 10 is realized by software, firmware, or a combination of software and firmware. The software or firmware is written as a program and stored in the memory 92. In the processing circuit 90, each function is realized by the processor 91 reading and executing the program stored in the memory 92. That is, the processing circuit 90 includes the memory 92 for storing the program that results in the processing of the automatic train operation system 10 being executed. It can also be said that these programs cause a computer to execute the procedures and methods of the automatic train operation system 10.

[0057] The above program includes a running state calculation step in which the running state calculation unit 11 calculates the current running state of the train 20, and a command value generation step in which the command value generation unit 14 calculates a command acceleration to follow the running plan information, converts the command acceleration into a notch command, and outputs it.In the command value generation step, the command value generation unit 14 extracts a target trajectory from the running plan information up to a specified time ahead from the time step of the current position of the train 20, sets constraints on the predicted running state and the command acceleration change amount when the command acceleration is changed, calculates the predicted running state within the range of the constraint conditions based on a dynamic characteristics model, uses the sum of the sum of the squares of the difference between the predicted running state and the target trajectory and the sum of the absolute values ​​of the command acceleration change amount as an evaluation function, calculates the command acceleration change amount that minimizes the evaluation value of the evaluation function within the range of the constraint conditions, and calculates the command acceleration using the command acceleration change amount.It can also be said that this is a program that causes the automatic train operation system 10 to execute the following.

[0058] Here, the processor 91 may be a CPU (Central Processing Unit), a processing device, an arithmetic device, a microprocessor, a microcomputer, or a DSP (Digital Signal Processor), etc. Furthermore, the memory 92 may be, for example, a non-volatile or volatile semiconductor memory such as a RAM (Random Access Memory), a ROM (Read Only Memory), a flash memory, an EPROM (Erasable Programmable ROM), or an EEPROM (registered trademark) (Electrically EPROM), a magnetic disk, a flexible disk, an optical disk, a compact disk, a minidisk, or a DVD (Digital Versatile Disc).

[0059] Fig. 16 is a diagram showing an example in which the processing circuitry 93 that realizes the automatic train operation system 10 according to the first embodiment is configured with dedicated hardware. When the processing circuitry 93 is configured with dedicated hardware, the processing circuitry 93 shown in Fig. 16 corresponds to, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or a combination thereof. Each function of the automatic train operation system 10 may be realized by the processing circuitry 93 separately for each function, or all functions may be realized collectively by the processing circuitry 93.

[0060] It is also possible to implement some of the functions of the automatic train operation system 10 using dedicated hardware and some using software or firmware. In this way, the processing circuit can implement each of the above-mentioned functions using dedicated hardware, software, firmware, or a combination of these.

[0061] As described above, according to this embodiment, in the automatic train operation system 10, the command value generation unit 14 extracts a target trajectory from the trip plan information stored in the trip plan information storage unit 12 for a time period specified from the time step of the current position of the train 20, and sets constraints on the predicted running state and the commanded acceleration change amount. The command value generation unit 14 calculates a predicted running state when the commanded acceleration is changed based on the dynamic characteristics model stored in the dynamic characteristics model storage unit 13, within the range of the constraints. The command value generation unit 14 uses the sum of the sum of the squares of the difference between the predicted running state and the target trajectory and the sum of the absolute values ​​of the commanded acceleration change amount as an evaluation function, calculates a commanded acceleration change amount that minimizes the evaluation value V of the evaluation function within the range of the constraints, and calculates the commanded acceleration using the commanded acceleration change amount.

[0062] As a result, the automatic train operation system 10 can reduce the number of notch switching times during tracking control by suppressing the variation in command acceleration even when the detection error becomes large, thereby improving the ride comfort of the train 20. Furthermore, the automatic train operation system 10 does not provide a dead zone, so there is no decrease in responsiveness. In this way, the automatic train operation system 10 can achieve both ride comfort for the train 20 and responsiveness of control for the train 20 while suppressing the effects of detection errors in the running state of the train 20.

[0063] Embodiment 2 In the second embodiment, a case where the conversion process of the command acceleration in the notch command conversion unit 145 is verified will be described.

[0064] Fig. 17 is a diagram showing an example of the configuration of the command value generation unit 14 included in the automatic train operation system 10 according to the second embodiment. Although not shown, the configuration of the automatic train operation system 10 according to the second embodiment is similar to the configuration of the automatic train operation system 10 according to the first embodiment shown in Fig. 1. The command value generation unit 14 according to the second embodiment additionally includes a verification unit 146 in addition to the command value generation unit 14 according to the first embodiment shown in Fig. 4.

[0065] In the second embodiment, the notch command conversion unit 145 performs conversion processing on the command acceleration acquired from the command acceleration calculation unit 144, as in the first embodiment, but converts the command acceleration into a notch and outputs the converted notch as a converted notch to the verification unit 146. Furthermore, the notch command conversion unit 145 outputs a converted command acceleration, which is a command acceleration equivalent to the converted notch, to the verification unit 146. Note that the converted command acceleration, which is a command acceleration equivalent to the converted notch, is not a continuous value but a quantized value according to the converted notch.

[0066] The verification unit 146 performs a verification operation using the converted notch and converted command acceleration acquired from the notch command conversion unit 145, and outputs the converted notch as a notch command or a modified converted notch as a notch command to the drive device 25 based on the result of the verification operation. Specifically, the verification unit 146 predicts the running state of the train 20 based on the converted command acceleration corresponding to the converted notch, and performs processing to lower the notch if it is determined that the train 20 will exceed the speed limit. The verification unit 146 acquires constraint conditions from the state constraint generation unit 142, acquires a command acceleration change amount from the command acceleration change amount calculation unit 143, and acquires a command acceleration from the command acceleration calculation unit 144. The verification unit 146 performs one of the following operations depending on whether the following verification conditions are met:

[0067] <Verification conditions> The verification conditions are when the command acceleration change amount is a negative value and it is determined that deceleration is necessary, and when the command acceleration is less than the converted command acceleration and it is determined that the conversion notch cannot reduce the acceleration to the optimum level. If these two requirements indicated by the verification conditions are met, the verification unit 146 performs an operation when the verification condition is met, and if at least one of the two requirements indicated by the verification conditions is not met, it performs an operation when the verification condition is not met.

[0068] <Action when verification conditions are met> The verification unit 146 determines that, as a result of the notch conversion, there is a possibility that the train 20 will exceed the speed limit or deviate from the constraints. Similar to the command acceleration change amount calculation unit 143, the verification unit 146 calculates a predicted running state when the converted command acceleration is output, assuming that the converted command acceleration will continue for L time steps. When the verification unit 146 determines that the speed value of the predicted running state will exceed the speed limit and deviate from the constraints, the verification unit 146 selects a notch that is changed from the converted notch in a direction that decreases the acceleration by a specified number of steps, and outputs the notch command to the drive device 25. Similarly to the notch command conversion unit 145, the verification unit 146 also inversely converts the acceleration value corresponding to the notch command to calculate an actual command acceleration, and outputs the actual command acceleration to the command acceleration change amount calculation unit 143 and the command acceleration calculation unit 144. When the verification unit 146 selects a notch that is changed from the converted notch in a direction that decreases the acceleration by a specified number of steps, the specified number of steps is, for example, one, but is not limited to this and can be two or more. If the speed value of the predicted traveling state is smaller than the speed limit, the verification unit 146 can also select a notch that is changed from the conversion notch in a direction that increases the acceleration by a specified number of steps.

[0069] <Action when verification conditions are not met> Outside the verification condition range, the verification unit 146 outputs the converted notch as a notch command to the drive device 25. Furthermore, the verification unit 146 outputs the converted command acceleration acquired from the notch command conversion unit 145 as an actual command acceleration to the command acceleration change amount calculation unit 143 and the command acceleration calculation unit 144.

[0070] Fig. 18 is a flowchart showing the operation of the command value generating unit 14 included in the automatic train operation system 10 according to the second embodiment. In Fig. 18, the operation from step S21 to step S24 is the same as the operation from step S21 to step S24 in the flowchart of the first embodiment shown in Fig. 14. The notch command converting unit 145 converts the command acceleration U k is converted into a notch and output to the verification unit 146 (step S31). The verification unit 146 outputs a notch command depending on whether the verification conditions are met, as described above (step S32).

[0071] The command value generation unit 14 calculates the optimal command acceleration that results in a predicted running state that follows the target trajectory within the range of constraints that take the speed limit into account, but the command acceleration that can actually be output is limited by the selectable notches. Therefore, as a result of converting to a notch, the predicted running state may differ from the assumption, and the speed limit may be exceeded. Therefore, in the second embodiment, the command value generation unit 14 re-predicts the running state based on the converted command acceleration corresponding to the converted notch, and lowers the notch if it determines that the speed limit will be exceeded. This allows the automatic train operation system 10 to prevent a situation in which the speed limit is exceeded due to a command output difference when converting from the optimal command acceleration to a notch.

[0072] The configurations shown in the above embodiments are merely examples, and may be combined with other known technologies, or different embodiments may be combined with each other. It is also possible to omit or modify parts of the configurations as long as they do not deviate from the gist of the invention.

[0073] Various aspects of the present disclosure are summarized below as appendices.

[0074] (Appendix 1) An automatic train operation system that controls the speed of a train by a notch command to a drive device that provides driving force to the train, a running state calculation unit that calculates a current running state of the train; a running plan information storage unit that stores running plan information between stations where the train runs; a dynamic characteristics model storage unit that stores a dynamic characteristics model of the drive device; a command value generation unit that calculates a command acceleration for following the travel plan information, converts the command acceleration into the notch command, and outputs the notch command; Equipped with The command value generation unit extracts a target trajectory from the running plan information up to a time ahead specified from the time step of the current position of the train, sets constraint conditions for a predicted running state and a command acceleration change amount when the command acceleration is changed, calculates the predicted running state within the range of the constraint conditions based on the dynamic characteristics model, uses the sum of the sum of squares of the difference between the predicted running state and the target trajectory and the sum of the absolute values ​​of the command acceleration change amount as an evaluation function, calculates the command acceleration change amount that minimizes the evaluation value of the evaluation function within the range of the constraint conditions, and calculates the command acceleration using the command acceleration change amount. An automatic train operation system characterized by: (Appendix 2) The command value generating unit a target trajectory extraction unit that extracts the target trajectory from the driving plan information; a state constraint generation unit that sets the constraint conditions; a command acceleration change amount calculation unit that calculates the predicted traveling state within the range of the constraint condition and calculates the command acceleration change amount that minimizes the evaluation value within the range of the constraint condition; a command acceleration calculation unit that calculates the command acceleration using the command acceleration change amount; a notch command conversion unit that converts the command acceleration into the notch command and outputs the notch command; 2. An automatic train operation system according to claim 1, comprising: (Appendix 3) The command value generating unit a target trajectory extraction unit that extracts the target trajectory from the driving plan information; a state constraint generation unit that sets the constraint conditions; a command acceleration change amount calculation unit that calculates the predicted traveling state within the range of the constraint condition and calculates the command acceleration change amount that minimizes the evaluation value within the range of the constraint condition; a command acceleration calculation unit that calculates the command acceleration using the command acceleration change amount; a notch command conversion unit that converts the command acceleration into a notch and outputs the converted notch after conversion and a converted command acceleration that is the command acceleration corresponding to the converted notch; a verification unit that performs a verification operation using the conversion notch and the conversion command acceleration, and outputs the conversion notch as the notch command or a modified version of the conversion notch as the notch command based on a result of the verification operation; 2. An automatic train operation system according to claim 1, comprising: (Appendix 4) when weighting the square of the difference between the predicted running state and the target trajectory at each time step, the command acceleration change amount calculation unit increases the weighting as the square of the difference between the predicted running state and the target trajectory at a time step closer to the present increases, 4. An automatic train operation system according to claim 2 or 3. (Appendix 5) the command acceleration change amount calculation unit weights the absolute value of the command acceleration change amount for each time step; 5. An automatic train operation system according to any one of appendices 2 to 4. (Appendix 6) the running state calculation unit calculates a running state including at least the current position and the current speed of the train using a speed signal indicating a value corresponding to the current speed of the train obtained from a tachograph and location information of the train obtained from information obtained by an on-board coil from a ground coil; 6. An automatic train operation system according to any one of appendices 1 to 5, (Appendix 7) The operation plan information storage unit stores, as the operation plan information, at least a running time, a target position, a target speed, and a speed limit between stations along which the train runs. 7. An automatic train operation system according to any one of appendices 1 to 6, (Appendix 8) 1. An automatic train operation method for an automatic train operation system that controls the speed of a train by a notch command to a drive device that provides driving force to the train, comprising: The automatic train operation system comprises: a running plan information storage unit that stores running plan information between stations where the train runs; a dynamic characteristics model storage unit that stores a dynamic characteristics model of the drive device; It is equipped with a running state calculation step in which a running state calculation unit calculates a current running state of the train; a command value generating step in which a command value generating unit calculates a command acceleration for following the driving plan information, converts the command acceleration into the notch command, and outputs the notch command; Including, In the command value generation step, the command value generation unit extracts a target trajectory from the running plan information up to a time ahead specified from the time step of the current position of the train, sets constraint conditions for a predicted running state and a command acceleration change amount when the command acceleration is changed, calculates the predicted running state within the range of the constraint conditions based on the dynamic characteristics model, sets the sum of the sum of squares of the difference between the predicted running state and the target trajectory and the sum of the absolute values ​​of the command acceleration change amount as an evaluation function, calculates the command acceleration change amount that minimizes the evaluation value of the evaluation function within the range of the constraint conditions, and calculates the command acceleration using the command acceleration change amount. An automatic train operation method characterized by the above. (Appendix 9) The command value generating step a target trajectory extraction step in which a target trajectory extraction unit extracts the target trajectory from the driving plan information; a state constraint generating step in which a state constraint generating unit sets the constraint conditions; a command acceleration change amount calculation step in which a command acceleration change amount calculation unit calculates the predicted traveling state within the range of the constraint condition and calculates the command acceleration change amount that minimizes the evaluation value within the range of the constraint condition; a command acceleration calculation step in which a command acceleration calculation unit calculates the command acceleration using the command acceleration change amount; a notch command converting step in which a notch command converting unit converts the command acceleration into the notch command and outputs the notch command; 9. The automatic train operation method according to claim 8, comprising: (Appendix 10) The command value generating step a target trajectory extraction step in which a target trajectory extraction unit extracts the target trajectory from the driving plan information; a state constraint generating step in which a state constraint generating unit sets the constraint conditions; a command acceleration change amount calculation step in which a command acceleration change amount calculation unit calculates the predicted traveling state within the range of the constraint condition and calculates the command acceleration change amount that minimizes the evaluation value within the range of the constraint condition; a command acceleration calculation step in which a command acceleration calculation unit calculates the command acceleration using the command acceleration change amount; a notch command converting step in which a notch command converting unit converts the command acceleration into a notch and outputs a converted notch after conversion and a converted command acceleration that is the command acceleration corresponding to the converted notch; a verification step in which a verification unit performs a verification operation using the conversion notch and the conversion command acceleration, and outputs the conversion notch as the notch command or a modified version of the conversion notch as the notch command based on a result of the verification operation; 9. The automatic train operation method according to claim 8, comprising: (Appendix 11) In the command acceleration change amount calculation step, when weighting the square of the difference between the predicted running state and the target trajectory at each time step, the command acceleration change amount calculation unit increases the weighting as the square of the difference between the predicted running state and the target trajectory at a time step closer to the present increases. 11. The automatic train operation method according to claim 9 or 10, (Appendix 12) In the command acceleration change amount calculation step, the command acceleration change amount calculation unit weights the absolute value of the command acceleration change amount for each time step. 12. An automatic train operation method according to any one of appendices 9 to 11, (Appendix 13) In the running state calculation step, the running state calculation unit calculates a running state including at least the current position and the current speed of the train using a speed signal indicating a value corresponding to the current speed of the train obtained from a tachograph and location information of the train obtained from information obtained by an on-board coil from a ground coil. 13. The automatic train operation method according to any one of appendices 8 to 12, (Appendix 14) The operation plan information storage unit stores, as the operation plan information, at least a running time, a target position, a target speed, and a speed limit between stations along which the train runs. 14. The automatic train operation method according to any one of appendices 8 to 13, [Explanation of symbols]

[0075] 10 Automatic train operation system, 11 Running state calculation unit, 12 Running plan information storage unit, 13 Dynamic characteristic model storage unit, 14 Command value generation unit, 20 Train, 21 Speed ​​generator, 22 On-board coil, 23 Power receiver, 24 ATC on-board equipment, 25 Drive unit, 31 Ground coil, 32 ATC ground equipment, 33 Rail, 90, 93 Processing circuit, 91 Processor, 92 Memory, 141 Target trajectory extraction unit, 142 State constraint generation unit, 143 Command acceleration change amount calculation unit, 144 Command acceleration calculation unit, 145 Notch command conversion unit, 146 Verification unit.

Claims

1. An automatic train operation system that controls the speed of a train by a notch command to a drive device that provides driving force to the train, a running state calculation unit that calculates a current running state of the train; a running plan information storage unit that stores running plan information between stations where the train runs; a dynamic characteristics model storage unit that stores a dynamic characteristics model of the drive device; a command value generation unit that calculates a command acceleration for following the travel plan information, converts the command acceleration into the notch command, and outputs the notch command; Equipped with The command value generation unit extracts a target trajectory from the running plan information up to a time ahead specified from the time step of the current position of the train, sets constraint conditions for a predicted running state and a command acceleration change amount when the command acceleration is changed, calculates the predicted running state within the range of the constraint conditions based on the dynamic characteristics model, uses the sum of the sum of squares of the difference between the predicted running state and the target trajectory and the sum of the absolute values ​​of the command acceleration change amount as an evaluation function, calculates the command acceleration change amount that minimizes the evaluation value of the evaluation function within the range of the constraint conditions, and calculates the command acceleration using the command acceleration change amount. An automatic train operation system characterized by:

2. The command value generating unit a target trajectory extraction unit that extracts the target trajectory from the driving plan information; a state constraint generation unit that sets the constraint conditions; a command acceleration change amount calculation unit that calculates the predicted traveling state within the range of the constraint condition and calculates the command acceleration change amount that minimizes the evaluation value within the range of the constraint condition; a command acceleration calculation unit that calculates the command acceleration using the command acceleration change amount; a notch command conversion unit that converts the command acceleration into the notch command and outputs the notch command; 2. The automatic train operation system according to claim 1, further comprising:

3. The command value generating unit a target trajectory extraction unit that extracts the target trajectory from the driving plan information; a state constraint generation unit that sets the constraint conditions; a command acceleration change amount calculation unit that calculates the predicted traveling state within the range of the constraint condition and calculates the command acceleration change amount that minimizes the evaluation value within the range of the constraint condition; a command acceleration calculation unit that calculates the command acceleration using the command acceleration change amount; a notch command conversion unit that converts the command acceleration into a notch and outputs the converted notch after conversion and a converted command acceleration that is the command acceleration corresponding to the converted notch; a verification unit that performs a verification operation using the conversion notch and the conversion command acceleration, and outputs the conversion notch as the notch command or a modified version of the conversion notch as the notch command based on a result of the verification operation; 2. The automatic train operation system according to claim 1, further comprising:

4. when weighting the square of the difference between the predicted running state and the target trajectory at each time step, the command acceleration change amount calculation unit increases the weighting as the square of the difference between the predicted running state and the target trajectory at a time step closer to the present increases, 4. An automatic train operation system according to claim 2 or 3.

5. the command acceleration change amount calculation unit weights the absolute value of the command acceleration change amount for each time step; 4. An automatic train operation system according to claim 2 or 3.

6. the running state calculation unit calculates a running state including at least the current position and the current speed of the train using a speed signal indicating a value corresponding to the current speed of the train obtained from a tachograph and location information of the train obtained from information obtained by an on-board coil from a ground coil; 4. An automatic train operation system according to claim 1, wherein:

7. The operation plan information storage unit stores, as the operation plan information, at least a running time, a target position, a target speed, and a speed limit between stations along which the train runs.

4. An automatic train operation system according to claim 1, wherein:

8. 1. An automatic train operation method for an automatic train operation system that controls the speed of a train by a notch command to a drive device that provides driving force to the train, comprising: The automatic train operation system comprises: a running plan information storage unit that stores running plan information between stations where the train runs; a dynamic characteristics model storage unit that stores a dynamic characteristics model of the drive device; It is equipped with a running state calculation step in which a running state calculation unit calculates a current running state of the train; a command value generating step in which a command value generating unit calculates a command acceleration for following the driving plan information, converts the command acceleration into the notch command, and outputs the notch command; Including, In the command value generation step, the command value generation unit extracts a target trajectory from the running plan information up to a time ahead specified from the time step of the current position of the train, sets constraint conditions for a predicted running state and a command acceleration change amount when the command acceleration is changed, calculates the predicted running state within the range of the constraint conditions based on the dynamic characteristics model, sets the sum of the sum of squares of the difference between the predicted running state and the target trajectory and the sum of the absolute values ​​of the command acceleration change amount as an evaluation function, calculates the command acceleration change amount that minimizes the evaluation value of the evaluation function within the range of the constraint conditions, and calculates the command acceleration using the command acceleration change amount. An automatic train operation method characterized by the above.

9. The command value generating step a target trajectory extraction step in which a target trajectory extraction unit extracts the target trajectory from the driving plan information; a state constraint generating step in which a state constraint generating unit sets the constraint conditions; a command acceleration change amount calculation step in which a command acceleration change amount calculation unit calculates the predicted traveling state within the range of the constraint condition and calculates the command acceleration change amount that minimizes the evaluation value within the range of the constraint condition; a command acceleration calculation step in which a command acceleration calculation unit calculates the command acceleration using the command acceleration change amount; a notch command converting step in which a notch command converting unit converts the command acceleration into the notch command and outputs the notch command; 9. The automatic train operation method according to claim 8, further comprising:

10. The command value generating step a target trajectory extraction step in which a target trajectory extraction unit extracts the target trajectory from the driving plan information; a state constraint generating step in which a state constraint generating unit sets the constraint conditions; a command acceleration change amount calculation step in which a command acceleration change amount calculation unit calculates the predicted traveling state within the range of the constraint condition and calculates the command acceleration change amount that minimizes the evaluation value within the range of the constraint condition; a command acceleration calculation step in which a command acceleration calculation unit calculates the command acceleration using the command acceleration change amount; a notch command converting step in which a notch command converting unit converts the command acceleration into a notch and outputs a converted notch after conversion and a converted command acceleration that is the command acceleration corresponding to the converted notch; a verification step in which a verification unit performs a verification operation using the conversion notch and the conversion command acceleration, and outputs the conversion notch as the notch command or a modified version of the conversion notch as the notch command based on a result of the verification operation; 9. The automatic train operation method according to claim 8, further comprising:

11. In the command acceleration change amount calculation step, when weighting the square of the difference between the predicted running state and the target trajectory at each time step, the command acceleration change amount calculation unit increases the weighting as the square of the difference between the predicted running state and the target trajectory at a time step closer to the present increases.

11. The automatic train operation method according to claim 9 or 10.

12. In the command acceleration change amount calculation step, the command acceleration change amount calculation unit weights the absolute value of the command acceleration change amount for each time step.

11. The automatic train operation method according to claim 9 or 10.

13. In the running state calculation step, the running state calculation unit calculates a running state including at least the current position and the current speed of the train using a speed signal indicating a value corresponding to the current speed of the train obtained from a tachograph and location information of the train obtained from information obtained by an on-board coil from a ground coil.

11. An automatic train operation method according to claim 8, wherein:

14. The operation plan information storage unit stores, as the operation plan information, at least a running time, a target position, a target speed, and a speed limit between stations along which the train runs.

11. An automatic train operation method according to claim 8, wherein:

Images