Response time evaluation method

The method optimizes smoothing intervals to accurately calculate response time in dynamometer systems by adjusting to noise levels, improving accuracy and reducing delays in signal evaluation.

JP2024164719AActive Publication Date: 2024-11-27MEIDENSHA CORP
View PDF 6 Cites 0 Cited by

Patent Information

Application Number
JP2023080401
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Filing Date
2023-05-15
Publication Date
2024-11-27
Estimated Expiration
2043-05-15

AI Technical Summary

Technical Problem

Existing methods for calculating the response time of evaluation signals in dynamometer systems are inaccurate when the signal-to-noise ratio is low due to excessive smoothing caused by median filter processing.

Method used

A method that calculates response time by optimizing the smoothing interval length based on the variance of the evaluation signal, using moving average and noise sample intervals to accurately determine the response time regardless of noise levels.

Benefits of technology

Enables precise calculation of response time with minimal delay, even in conditions with varying noise levels, ensuring accurate evaluation of dynamometer system performance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2024164719000001_ABST
    Figure 2024164719000001_ABST
Patent Text Reader

Abstract

To provide a response time evaluation method capable of precisely evaluating response time of an evaluation signal by performing processing corresponding to the level of noise included in the original evaluation signal.SOLUTION: A response time evaluation method includes the steps of: acquiring time-series data x(i) of an evaluation signal outputted from a control object when a step input signal is inputted to the control object; calculating unbiased variance σn2 of time-series data x(i) for a noise sample interval with a length N determined within a stationary interval of a step input signal; calculating time-series data σnp2(i), xma(i) of sample dispersion and moving average of time-series data x(i) for a smoothing interval with a length of NP and with discrete time (I) as a center; calculating time-series data y(i) of a smoothing signal on the basis of time-series data σnp2(i), xma(i); and calculating response time of an evaluation signal on the basis of time-series data y(i).SELECTED DRAWING: Figure 3
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to a response time evaluation method, and more particularly to a response time evaluation method for evaluating the response time of an evaluation signal output from a controlled object or a dynamometer system. [Background technology]

[0002] A chassis dynamometer system is used in tests to measure and evaluate a vehicle's power consumption rate, fuel consumption rate, exhaust purification performance, etc. The chassis dynamometer system includes rollers on which the wheels of the vehicle to be tested are placed and a dynamometer connected to the rollers, and reproduces conditions close to those of actual driving by applying running resistance, such as rolling resistance and inertial resistance, to the vehicle running on the rollers using the dynamometer and rollers.

[0003] In addition, in tests using such chassis dynamometer systems, it is necessary to evaluate whether the response time of the dynamometer is appropriate in order to guarantee the measurement and evaluation results of the test. The response time is often calculated by measuring the change in a predetermined evaluation signal output from a controlled object when a step-changing input signal is input to the controlled object. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2022-148844 A Summary of the Invention [Problem to be solved by the invention]

[0005] Furthermore, in many cases, the evaluation signal output from the sensor contains noise, and therefore the technology described in Patent Document 1 applies median filtering to the time series data of the evaluation signal to remove the noise contained in the original evaluation signal while preventing a delay in the rising edge of the evaluation signal. More specifically, the technology described in Patent Document 1 optimizes the time width of the median filter so that the signal that has been subjected to median filtering falls within a range between a predetermined upper and lower limits.

[0006] Therefore, according to the technology described in Patent Document 1, if the S / N ratio of the evaluation signal is large, the response time of the evaluation signal can be calculated with high accuracy. However, if the S / N ratio is small, the signal may be excessively smoothed by the median filter process, which may make it impossible to calculate the response time with high accuracy.

[0007] An object of the present invention is to provide a response time evaluation method capable of evaluating the response time of an evaluation signal with high accuracy by performing processing according to the magnitude of noise contained in the original evaluation signal. [Means for solving the problem]

[0008] (1) A response time evaluation method according to the present invention is a method for evaluating a response time of an evaluation signal output from a controlled object, comprising the steps of: (A) acquiring time series data x(i) (hereinafter, i is a parameter representing a discrete time) of the evaluation signal output from the controlled object when a step input signal that changes stepwise is input to the controlled object; (B) calculating a variance σ of the time series data x(i) over a noise sampling interval defined within a stationary interval of the step input signal; n 2 (C) calculating time series data σ of the variance of the time series data x(i) over a smoothing interval centered on the discrete time i and having a length equal to or less than the noise sample interval. np 2 (i) and the time series data of the moving average x ma(i) calculating time series data y(i) of a smoothed signal defined by the following equation (1); and (D) calculating a response time of the evaluation signal based on the time series data y(i).

number

[0009] (2) In this case, in the step (D), it is preferable to calculate, as the response time, a time from a rising edge of the step input signal to a time when the time series data y(i) exceeds a threshold value determined based on the step input signal.

[0010] (3) In this case, it is preferable that the steady section is a section before or after the step input signal changes stepwise.

[0011] (4) In this case, it is preferable that the controlled object includes a rotating body that rotates in response to the input signal.

[0012] (5) In this case, in the step (C), the time series data σ np 2 (i) Peak value σ in the steady state section peak 2 is the variance σ n 2 It is preferable to optimize the smoothing interval length so that the smoothing interval falls within a variance setting range that is determined to include the above-mentioned.

[0013] (6) In this case, in the step (C), the time series data σ np 2 (i) Peak value σ in the steady state section peak 2 is the variance σ n 2It is preferable to optimize a smoothing interval length, which is the length of the smoothing interval, so that the time series data y(i) in the steady interval falls within a variance setting range that is determined to include a command value corresponding to the step input signal, and the time series data y(i) in the steady interval falls within a signal setting range that is determined to include a command value corresponding to the step input signal.

[0014] (7) A response time evaluation method according to the present invention is a method for evaluating a response time of an evaluation signal in a dynamometer system including a dynamometer, an inverter that supplies power to the dynamometer in response to an input signal, and an evaluation signal output unit that outputs an evaluation signal in response to a speed or torque of the dynamometer, the method comprising the steps of: (A) acquiring time series data x(i) (hereinafter, i is a parameter representing a discrete time) of the evaluation signal output from the sensor when a step input signal that changes stepwise is input to the inverter; (B) calculating a variance σ of the time series data x(i) over a noise sampling interval defined within a stationary interval of the step input signal; n 2 (C) calculating time series data σ of the variance of the time series data x(i) over a smoothing interval centered on the discrete time i and having a length equal to or less than the noise sample interval. np 2 (i) and the time series data of the moving average x ma (i) calculating time series data y(i) of a smoothed signal defined by the following equation (2); and (D) calculating a response time of the evaluation signal based on the time series data y(i).

number

[0015] (1) In the present invention, based on the time series data x(i) of the evaluation signal, the variance σ of the time series data x(i) over a noise sample interval defined within the stationary interval of the step input signal is calculated. n 2and the time series data σ of the variance of the time series data x(i) over a smoothing interval centered on the discrete time i and having a length equal to or less than the noise sample interval. np 2 (i) and the time series data of the moving average x ma (i) is calculated. In the present invention, the time series data y(i) of the smoothed signal obtained by smoothing the evaluation signal based on the above formula (1) is calculated, and the response time of the original evaluation signal is calculated based on this time series data y(i). According to the above formula (1), when the change in the evaluation signal is small in the smoothing section centered on the discrete time i, that is, when most of the fluctuation components of the evaluation signal are noise, σ np 2 (i) ≒ σ n 2 Therefore, y(i)≒x ma (i). In other words, in the smoothing section where the change in the evaluation signal is small, the noise contained in the original evaluation signal can be removed, and smooth time series data y(i) can be obtained. On the other hand, when the change in the evaluation signal is large in the smoothing section centered on the discrete time i, that is, when the evaluation signal changes in a step-like manner, σ np 2 (i)>σ n 2 Therefore, y(i) ≈ x(i). That is, in the smoothing section where the change in the evaluation signal is large, it is possible to obtain time series data y(i) that is approximately equal to the original time series data x(i). Therefore, according to the present invention, when generating the time series data y(i) of the smoothed signal from the time series data x(i) of the original evaluation signal, the fluctuation due to noise in the stationary section is smoothed, and no delay due to the smoothing process occurs in the section where the time series data x(i) changes stepwise, so that the response time of the evaluation signal can be calculated with high accuracy.

[0016] (2) In the present invention, the response time is calculated as the time from the rising edge of the step input signal to the time when the time series data y(i) of the smoothed signal exceeds a threshold value determined based on the step input signal. This makes it possible to accurately calculate the response time of the evaluation signal to the step input signal in the controlled object.

[0017] (3) In the present invention, the interval before or after the step input signal changes stepwise, i.e., the interval in which most of the fluctuation components of the evaluation signal are noise, is defined as a stationary interval, and a noise sample interval is defined within this stationary interval. The variance σ of the time series data x(i) is calculated. n 2 This makes it possible to perform an appropriate smoothing process (i.e., the process according to the above formula (1)) according to the level of noise contained in the evaluation signal.

[0018] (4) In the present invention, the response time of an evaluation signal to a step input signal in a controlled object having a rotating body can be calculated with high accuracy.

[0019] (5) In the present invention, the time series data σ np 2 Peak value σ in the steady state section of (i) peak 2 is the variance σ n 2 Therefore, according to the present invention, the smoothing interval length can be set to a length according to the magnitude of noise contained in the evaluation signal.

[0020] (6) In the present invention, the time series data σ np 2 Peak value σ in the steady state section of (i) peak 2 is the variance σ n 2 The smoothing interval length is optimized so that the time series data y(i) in the steady interval falls within a variance setting range determined to include the command value corresponding to the step input signal, and the time series data y(i) in the steady interval falls within a signal setting range determined to include the command value corresponding to the step input signal. Thus, according to the present invention, the smoothing interval length is set to the minimum required length, and the fluctuations in the time series data y(i) in the steady interval can be kept within the signal setting range, resulting in accurate response time evaluation.

[0021] (7) According to the present invention, for the same reason as in the invention (1) above, the response time of an evaluation signal in a dynamometer system can be evaluated (that is, calculated) with high accuracy. [Brief description of the drawings]

[0022] [Figure 1] 1 is a diagram showing a configuration of a chassis dynamometer system to which a response time evaluation method according to a first embodiment of the present invention is applied. [Figure 2A] 6 is a flowchart (part 1) showing a specific procedure of the response time evaluation process according to the embodiment. [Figure 2B] 6 is a flowchart showing a specific procedure of the response time evaluation process according to the embodiment (part 2). [Diagram 3] 4 is a time chart for explaining a procedure of a response time evaluation process according to the embodiment. [Figure 4A] 10 is a flowchart showing a specific procedure of a response time evaluation process according to the second embodiment of the present invention (part 1). [Figure 4B] 6 is a flowchart showing a specific procedure of the response time evaluation process according to the embodiment (part 2). [Diagram 5] 4 is a time chart for explaining a procedure of a response time evaluation process according to the embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0023] First Embodiment Hereinafter, a first embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a diagram showing the configuration of a chassis dynamometer system S to which a response time evaluation method according to this embodiment is applied, and a test vehicle V (hereinafter simply referred to as "vehicle V").

[0024] In the following description, the vehicle V is a four-wheel drive (4WD) vehicle that transmits power separately to the front wheels Wf and the rear wheels Wr, but the present invention is not limited to this. The vehicle V may be a front-wheel drive (FWD) vehicle or a rear-wheel drive (RWD) vehicle.

[0025] The chassis dynamometer system S comprises front wheel rollers 1f and rear wheel rollers 1r on which the front wheels Wf and rear wheels Wr of the vehicle V can be placed, respectively, and which rotate in accordance with the rotation of the wheels, a front wheel dynamometer 2f and a rear wheel dynamometer 2r coaxially connected to the rollers 1f, 1r, respectively, a front wheel speed sensor 3f and a rear wheel speed sensor 3r which detect the rotational speed of the dynamometers 2f, 2r or the rollers 1f, 1r, respectively, a front wheel torque sensor 4f and a rear wheel torque sensor 4r which detect the torque acting on the dynamometers 2f, 2r, respectively, a front wheel inverter 5f and a rear wheel inverter 5r which supply power to the dynamometers 2f, 2r, respectively, a control device 6 which controls the dynamometers 2f, 2r by operating the inverters 5f, 5r, and an operation measurement system 7 connected to the control device 6.

[0026] The control device 6 performs speed control and running resistance control of the dynamometers 2f, 2r based on detection signals transmitted from the speed sensors 3f, 3r and torque sensors 4f, 4r, etc., and command signals transmitted from an operation measurement system 7 described below.

[0027] The operation measurement system 7 is a computer that has an operation reception function that receives various operations by an operator of the chassis dynamometer system S, a communication function that sends and receives various information to the control device 6, a calculation function that performs calculations based on information obtained by the operation reception function, communication function, etc., and a display function that displays the information obtained by the calculation function, communication function, etc. in a manner that can be recognized by the operator.

[0028] 2A and 2B are flowcharts showing a specific procedure of the response time evaluation process by the operation measurement system 7. This response time evaluation process is a process in which the above-mentioned dynamometer system S is treated as a control object, and a response time of an evaluation signal output from the control object when an input signal is input to the control object is evaluated.

[0029] In the following, the front wheel control system of the dynamometer system S described above, which is composed of the front wheel dynamometer 2f, the front wheel speed sensor 3f, the front wheel torque sensor 4f, and the front wheel inverter 5f, is the object of control, the torque current command signal input to the front wheel inverter 5f is the input signal, a signal obtained by performing a predetermined process on the speed detection signal output from the front wheel speed sensor 3f (for example, an acceleration / deceleration signal obtained by differentiating the speed detection signal) is the evaluation signal, and a case will be described in which the response time of this evaluation signal is evaluated, but the present invention is not limited to this. In such a front wheel control system of the dynamometer system S, the evaluation signal may be the torque detection signal output from the front wheel torque sensor 4f or a signal generated based on at least one of the speed detection signal and the torque detection signal (for example, the speed detection signal itself, or a driving force signal obtained by calculation using the speed detection signal and the torque detection signal).

[0030] The present invention can also be applied to a case where a rear wheel control system consisting of a rear wheel dynamometer 2r, a rear wheel speed sensor 3r, a rear wheel torque sensor 4r, and a rear wheel inverter 5r is the control target, a torque current command signal input to the rear wheel inverter 5r is the input signal, a signal obtained by performing a predetermined process on the speed detection signal output from the rear wheel speed sensor 3r (for example, an acceleration / deceleration signal obtained by differentiating the speed detection signal) is the evaluation signal, and the response time of this evaluation signal is evaluated. In this case, as with the front wheel control system, the torque detection signal output from the rear wheel torque sensor 4r or a signal generated based on at least one of the speed detection signal and the torque detection signal (for example, the speed detection signal itself, or a driving force signal obtained by calculation using the speed detection signal and the torque detection signal) may be the evaluation signal.

[0031] In the following, a case will be described in which the operation measurement system 7 starts the response time evaluation process shown in Figures 2A and 2B in response to receiving a predetermined start operation by the operator, but the present invention is not limited to this. The response time evaluation process shown in Figures 2A and 2B may be automatically started by the operation measurement system 7 during or before or after driving a vehicle test on a vehicle under the control of a control device. In this case, the result of the response time evaluation process can be stored in a storage medium (not shown) in association with the result of the test on the vehicle, or reported to the operator.

[0032] 3 is a time chart for explaining the procedure of the response time evaluation process. In the following, the specific procedure of the response time evaluation process shown in FIGS. 2A and 2B will be explained with reference to the time chart in FIG.

[0033] First, in step ST1, the operation measurement system 7 acquires time series data x(i) (hereinafter, i is an integer parameter representing a discrete time) of an evaluation signal output from a controlled object when a step input signal that changes stepwise is input to the controlled object, at a predetermined sampling period. More specifically, the operation measurement system 7 inputs a step input signal to the front wheel inverter 5r via the control device 6, and acquires time series data x(i) of the evaluation signal generated based on a speed detection signal output from the front wheel speed sensor 2f at this time. In FIG. 3, the time series data z(i) of the step input signal is indicated by a thick solid line, and the time series data x(i) of the evaluation signal is indicated by a thin solid line.

[0034] Next, in step ST2, the operation measurement system 7 determines a noise sample section of a predetermined length within the stationary section of the step input signal, and calculates an unbiased variance σ of the time series data x(i) over the noise sample section based on the time series data x(i) acquired in step ST1. n 2 Then, the process proceeds to step ST3. In the following, the noise sample interval length, which is the length of the noise sample interval, is represented by an integer value "N" that corresponds to the number of data items of the time series data x(i) included in this noise sample interval.

[0035] Here, the steady section of the step input signal is a section in which the time series data z(i) of the step input signal is almost constant, and is divided into a pre-change steady section before time ts when the step input signal changes stepwise, and a post-change steady section after time ts, as shown in Fig. 3. Note that, as shown in Fig. 3, the time series data x(i) of the evaluation signal starts to rise a predetermined time after time ts, so the start of the post-change steady section is set a predetermined time after ts.

[0036] In the following, we use the unbiased variance σ n 2 A case will be described in which the noise sampling interval for calculating is set within the pre-change stationary interval as shown in Fig. 3, but the present invention is not limited to this. The noise sampling interval is not limited to the pre-change stationary interval, and may be set within the post-change stationary interval.

[0037] The noise sample interval is set to be as wide as possible within the stationary interval, and the noise sample interval length N is therefore set to be slightly shorter than the length of the stationary interval, for example.

[0038] Next, in step ST3, the operation measurement system 7 sets the smoothing interval length NP, which is the length of the smoothing interval described later, to an initial value NP0 (NP←NP0). Here, the initial value NP0 is an arbitrary integer value, but it is preferable to set it to a value sufficiently smaller than the value of the noise sample interval length N described above (N>NP0).

[0039] Next, in step ST4, the operation measurement system 7 defines a smoothing interval in a range having a smoothing interval length NP centered on the discrete time i, and calculates time series data σ np 2 (i) is calculated, and the process proceeds to step ST5.

[0040] Next, in step ST5, the operation measurement system 7 calculates the time series data σ of the sample variance calculated in step ST4. np 2In the stationary section of (i) (more specifically, in the stationary section that is the same as the section in which the noise sample section was defined in step ST2, in the example of FIG. 3, in the stationary section before the change), the peak value σ peak 2 is the unbiased variance σ n 2 It is determined whether the dispersion setting range is within a range that is determined to include "Lower" and "Upper". More specifically, the dispersion setting range is defined by the following formula (3). In the following formula (3), "Lower" and "Upper" are both arbitrary values ​​slightly larger than 0.

number

[0041] If the determination result in step ST5 is NO, the operation measurement system 7 proceeds to step ST6, and if the determination result is YES, the operation measurement system 7 proceeds to step ST11 (see FIG. 2B).

[0042] Next, in step ST6, the operation measurement system 7 increases the smoothing section length NP by a predetermined unit length ΔNP, and then proceeds to step ST7 (NP←NP+ΔNP). Here, the value of the unit length ΔNP is an arbitrary integer value larger than 0, but is set to the smallest possible value, for example, "1."

[0043] Next, in step ST7, the manipulation measurement system 7 judges whether the current smoothing interval length NP is equal to or less than a predetermined upper limit length NPmax. Below, a case where the upper limit length NPmax is equal to the noise sample interval length N (NPmax=N) will be described, but the present invention is not limited to this. The upper limit length NPmax is set within a range that is greater than the initial value NP0 and less than or equal to the noise sample interval length N. If the judgment result in step ST7 is YES, the manipulation measurement system 7 returns to step ST4 and again measures the time series data σ of the sample variance under the changed smoothing interval. np 2 If (i) is calculated and the result is NO, the process proceeds to step ST11 (see FIG. 2B).

[0044] As described above, the operation measurement system 7 calculates the time series data σnp 2 Peak value σ in the steady state section of (i) peak 2 falls within the dispersion setting range shown in the above formula (3) (see step ST5) or the smoothing section length NP reaches the upper limit length NPmax (see step ST7). In other words, the operation measurement system 7 increases the smoothing section length NP from the initial value by unit length ΔNP. peak 2 The smoothing interval length NP is optimized so that falls within the variance setting range.

[0045] Next, in step ST11, the operation measurement system 7 calculates time series data x(i) of the moving average of the time series data x(i) over the smoothing interval optimized by the above-mentioned procedure. ma (i) is calculated, and the process proceeds to step ST12.

[0046] Next, in step ST12, the operation measurement system 7 calculates the unbiased variance σ n 2 and the time series data σ calculated under the optimized smoothing interval. np 2 (i),x ma The time-series data y(i) of the smoothed signal obtained by smoothing the original evaluation signal is calculated by the following equation (4) using (i) and (i), and the process proceeds to step ST13.

number

[0047] In Fig. 3, the time series data y(i) of the smoothed signal calculated by the above procedure is shown by a thick dashed line. According to the above formula (4), when the fluctuation of the evaluation signal is small in the smoothing section centered on the discrete time i, that is, when most of the fluctuation components of the evaluation signal are noise, σ np 2 (i) ≒ σ n 2 Therefore, y(i)≒x ma(i). That is, in the section where the change in the evaluation signal is small (i.e., the section excluding the vicinity of time ts in FIG. 3), the noise contained in the original evaluation signal can be removed to obtain smooth time series data y(i). On the other hand, when the change in the evaluation signal is large in the smoothing section centered on the discrete time i, that is, when the evaluation signal changes in a step-like manner, σ np 2 (i)>σ n 2 Therefore, y(i) ≈ x(i). That is, in a section where the change in the evaluation signal is large (i.e., a section near time ts in FIG. 3), it is possible to obtain time series data y(i) that is approximately equal to the original time series data x(i). Therefore, according to this embodiment, when generating the time series data y(i) of the smoothed signal from the time series data x(i) of the original evaluation signal, the fluctuation due to noise in the stationary section is smoothed, and no delay due to the smoothing process occurs in the section where the time series data x(i) changes stepwise.

[0048] Next, in step ST13, the operation measurement system 7 calculates the response time of the original evaluation signal based on the time series data y(i) of the smoothed signal calculated in step ST12, and ends the response time evaluation process shown in Figures 2A and 2B. More specifically, the operation measurement system 7 calculates, as the response time, the time from the rising time of the step input signal (time ts in Figure 3) to the time when the time series data y(i) of the smoothed signal exceeds a threshold value determined based on the step input signal.

[0049] The response time evaluation method according to the present embodiment provides the following advantages. (1) In this embodiment, based on the time series data x(i) of the evaluation signal, the unbiased variance σ of the time series data x(i) over the noise sample interval defined within the stationary interval of the step input signal is calculated. n 2 and the time series data σ of the sample variance of the time series data x(i) centered on the discrete time i and spanning a smoothing interval whose length is equal to or shorter than the noise sample interval. np 2 (i) and the time series data of the moving average x ma(i) is calculated. In this embodiment, time series data y(i) of a smoothed signal obtained by smoothing the evaluation signal based on the above formula (4) is calculated, and the response time of the original evaluation signal is calculated based on this time series data y(i). Thus, according to this embodiment, when generating the time series data y(i) of the smoothed signal from the time series data x(i) of the original evaluation signal, fluctuations due to noise in the steady interval are smoothed, and no delay due to the smoothing process occurs in the interval where the time series data x(i) changes stepwise, so that the response time of the evaluation signal can be calculated with high accuracy.

[0050] (2) In this embodiment, the response time is calculated as the time from the rising edge of the step input signal to the time when the time series data y(i) of the smoothed signal exceeds a threshold value determined based on the step input signal. This makes it possible to accurately calculate the response time of the evaluation signal to the step input signal in the controlled object.

[0051] (3) In this embodiment, the interval before or after the step input signal changes stepwise, i.e., the interval in which most of the fluctuation components of the evaluation signal are noise, is defined as a stationary interval, and a noise sample interval is defined within this stationary interval. The unbiased variance σ of the time series data x(i) is calculated. n 2 This makes it possible to perform an appropriate smoothing process (i.e., the process using the above formula (4)) according to the level of noise contained in the evaluation signal.

[0052] (4) In this embodiment, the time series data of the sample variance σ np 2 Peak value σ in the steady state section of (i) peak 2 is the unbiased variance σ n 2 Therefore, according to this embodiment, the smoothing section length NP can be set to a length according to the magnitude of the noise contained in the evaluation signal.

[0053] (5) According to this embodiment, the response time of the dynamometer system can be automatically evaluated not only during maintenance of the dynamometer system, but also during, before, and after vehicle test operation as long as the dynamometer is in a controlled state, thereby improving user convenience.

[0054] <Second embodiment> A second embodiment of the present invention will be described below with reference to the drawings. The mechanical configuration of the chassis dynamometer system according to this embodiment is the same as that of the chassis dynamometer system S according to the first embodiment, and therefore illustrations and detailed description are omitted. The chassis dynamometer system according to this embodiment differs from the chassis dynamometer system S according to the first embodiment in the procedure of response time evaluation processing in the operation measurement system 7.

[0055] 4A and 4B are flow charts showing specific steps of the response time evaluation process by the operation measurement system 7 according to this embodiment. In the following, as in the first embodiment, a case will be described in which the front wheel control system is the control target, the torque current command signal input to the front wheel inverter is the input signal, a signal obtained by performing a predetermined process on the speed detection signal output from the front wheel speed sensor (for example, an acceleration / deceleration signal obtained by differentiating the speed detection signal) is the evaluation signal, and the response time of this evaluation signal is evaluated; however, the present invention is not limited to this. The control target, the evaluation signal, etc. can be changed in the same way as in the first embodiment.

[0056] In the following, a case will be described in which the operation measurement system 7 starts the response time evaluation process shown in Figures 4A and 4B in response to receiving a predetermined start operation by the operator, as in the first embodiment, but the present invention is not limited to this. The response time evaluation process shown in Figures 4A and 4B may be automatically started by the operation measurement system 7 during or before or after driving a vehicle test on a vehicle under the control of a control device, as in the first embodiment.

[0057] Fig. 5 is a time chart for explaining the procedure of the response time evaluation process according to this embodiment. In the following, the specific procedure of the response time evaluation process shown in Fig. 4A and Fig. 4B will be explained with reference to the time chart in Fig. 5.

[0058] 4A are the same as steps ST1 to ST7 shown in FIG. 2A, and therefore detailed description thereof will be omitted below. That is, the operation measurement system 7 repeatedly executes steps ST24 to ST27 to obtain a peak value σ peak 2 After adjusting the smoothing section length NP so that falls within the variance setting range shown in the above formula (3), the process proceeds to step ST31 (see FIG. 4B).

[0059] In step ST31, the operation measurement system 7 defines a smoothing interval in a range having a smoothing interval length NP centered on the discrete time i, and calculates time series data σ(i) of the sample variance of the time series data x(i) over this smoothing interval. np 2 (i) and the time series data of the moving average x ma (i) is calculated, and the process proceeds to step ST32.

[0060] Next, in step ST32, the operation measurement system 7 calculates the unbiased variance σ n 2 and the time series data σ calculated in step ST31. np 2 (i),x ma Using (i), the time series data y(i) of the smoothed signal obtained by smoothing the original evaluation signal is calculated according to the above equation (4), and the process proceeds to step ST33.

[0061] Next, in step ST33, the operation measurement system 7 determines whether the time-series data y(i) calculated in step ST32 falls within a signal setting range determined to include the command value Sig_cmd corresponding to the step input signal within the pre-change steady section. More specifically, the signal setting range is defined by the following formula (5). In the following formula (5), "Sig_Lower" is set to a value slightly smaller than the command value Sig_cmd, and "Sig_Upper" is set to a value slightly larger than the command value Sig_cmd.

number

[0062] In this embodiment, a case will be described in which it is determined whether the time series data y(i) falls within the signal setting range only in the pre-change stationary section out of the two stationary sections, but the present invention is not limited to this. It may be determined whether the time series data y(i) falls within the signal setting range only in the post-change stationary section or in both the pre-change stationary section and the post-change stationary section.

[0063] If the determination result in step ST33 is NO, the operation measurement system 7 proceeds to step ST34, and if the determination result is YES, the operation measurement system 7 proceeds to step ST41.

[0064] Next, in step ST34, the operation measurement system 7 increases the smoothing section length NP by a predetermined unit length ΔNP, and then proceeds to step ST35 (NP←NP+ΔNP).

[0065] Next, in step ST35, the operation measurement system 7 judges whether or not the current smoothing section length NP is equal to or less than the upper limit length NPmax. If the judgment result in step ST35 is YES, the operation measurement system 7 returns to step ST31 and again measures the time series data σ np 2 (i),x ma (i), y(i) are calculated, and if the result is NO, the process proceeds to step ST41.

[0066] As described above, the operation measurement system 7 according to this embodiment uses the time series data σ np 2 Peak value σ in the steady state section of (i) peak 2 falls within the variance setting range shown in the above formula (3) (see step ST25) or the smoothing section length NP reaches the upper limit length NPmax (see step ST27), the smoothing section length NP is increased from the initial value by unit length ΔNP, and then the smoothing section length NP is further increased by unit length ΔNP until the time series data y(i) in the steady section falls within the signal setting range shown in the above formula (5) (see step ST33) or the smoothing section NP reaches the upper limit length NPmax (see step ST35). That is, the operation measurement system 7 repeatedly executes the processes of steps ST24 to ST27 and steps ST31 to ST35 to obtain the peak value σ peak 2 The smoothing interval length NP is optimized so that falls within a variance setting range and the time series data y(i) in the stationary interval falls within a signal setting range.

[0067] Next, in step ST41, the operation measurement system 7 calculates the response time of the original evaluation signal by the same procedure as in the first embodiment, based on the time-series data y(i) of the smoothed signal calculated under the smoothing section optimized by the procedure described above, and terminates the response time evaluation process shown in FIGS. 4A and 4B.

[0068] In addition to the above-mentioned advantages (1) to (3) and (5), the response time evaluation method according to this embodiment has the following advantages.

[0069] (6) In this embodiment, the time series data of the sample variance σ np 2 Peak value σ in the steady state section of (i) peak 2 is the unbiased variance σ n 2The smoothing section length NP is optimized so that the time series data y(i) in the steady section falls within a variance setting range that is determined to include the command value corresponding to the step input signal, and the time series data y(i) in the steady section falls within a signal setting range that is determined to include the command value corresponding to the step input signal. Therefore, according to this embodiment, the smoothing section length NP is set to the minimum required length, and the fluctuation of the time series data y(i) in the steady section can be kept within the signal setting range, resulting in the response time being evaluated with high accuracy.

[0070] Although one embodiment of the present invention has been described above, the present invention is not limited to this embodiment. The detailed configuration may be appropriately changed within the scope of the spirit of the present invention.

[0071] For example, in the above embodiment, the response time evaluation method is described as being applied to a chassis dynamometer system, but the present invention is not limited to this. The present invention can be applied not only to chassis dynamometer systems, but also to dynamometer systems such as engine bench systems and drive train bench systems, and to controlled objects having rotating bodies such as dynamometers and rotating electric machines. [Explanation of symbols]

[0072] S...Chassis dynamometer system (dynamometer system) 1f, 1r...Rola 2f, 2r...Dynamometer 3f, 3r...Speed ​​sensor 4f, 4r...Torque sensor 5f, 5r...Inverter 6...Control device 7. Operation and measurement system V…Vehicle

Claims

1. A method for evaluating a response time of an evaluation signal output from a controlled object, comprising the steps of: (A) acquiring time series data x(i) (hereinafter, i is a parameter representing a discrete time) of the evaluation signal output from the control target when a step input signal that changes stepwise is input to the control target; (B) the variance σ of the time series data x(i) over a noise sample interval defined within the stationary interval of the step input signal n 2 and calculating (C) time series data σ of the variance of the time series data x(i) over a smoothing interval centered on the discrete time i and having a length equal to or less than the noise sample interval np 2 (i) and the moving average time series data x ma (i) and calculating time series data y(i) of a smoothed signal defined by the following equation (1); (D) calculating a response time of the evaluation signal based on the time series data y(i). [0010]

2. 2. The response time evaluation method according to claim 1, wherein in the step (D), a time from a rising edge time of the step input signal to a time when the time series data y(i) exceeds a threshold value determined based on the step input signal is calculated as the response time.

3. 2. The response time evaluation method according to claim 1, wherein the steady-state section is a section before or after the step input signal changes stepwise.

4. 2. The response time evaluation method according to claim 1, wherein the controlled object comprises a rotating body that rotates in response to the input signal.

5. In the step (C), the time series data σ np 2 (i) Peak value σ in the steady state section peak 2 is the variance σ n 2 5. The response time evaluation method according to claim 1, further comprising optimizing a smoothing interval length, which is a length of the smoothing interval, so as to fall within a variance setting range determined to include:

6. In the step (C), the time series data σ np 2 (i) Peak value σ in the steady state section peak 2 is the variance σ n 2 5. The response time evaluation method according to claim 1 , further comprising the step of optimizing a smoothing interval length, which is a length of the smoothing interval, so that the time series data y(i) in the steady interval falls within a variance setting range determined to include a command value corresponding to the step input signal, and the time series data y(i) in the steady interval falls within a signal setting range determined to include a command value corresponding to the step input signal.

7. A dynamometer; an inverter for supplying power to the dynamometer in response to an input signal; an evaluation signal output unit that outputs an evaluation signal corresponding to a speed or a torque of the dynamometer, (A) acquiring time series data x(i) (hereinafter, i is a parameter representing a discrete time) of the evaluation signal output from the sensor when a step input signal that changes stepwise is input to the inverter; (B) the variance σ of the time series data x(i) over a noise sample interval defined within the stationary interval of the step input signal n 2 and calculating (C) time series data σ of the variance of the time series data x(i) over a smoothing interval centered on the discrete time i and having a length equal to or less than the noise sample interval np 2 (i) and the moving average time series data x ma(i) and calculating time series data y(i) of a smoothed signal defined by the following equation (2); (D) calculating a response time of the evaluation signal based on the time series data y(i). [0025]

Citation Information

Patent Citations

  • Speed matching control apparatus

    JP1986053543A

  • Smoothing processing method, noise removal device by smoothing processing, light pulse testing device using same

    JP1992069541A

  • Electrical inertia evaluation device for dynamometer and its method

    JP2004219185A

  • Performance evaluation method and system for chassis dynamometer

    JP2006090795A

  • Dynamometer control device

    JP2020139854A