Sensor output delay calibration device and method

Abstract

The present invention relates to a sensor output delay calibration device and method, wherein a synchronous signal generating device in the calibration device is used to generate a first signal and a second signal for synchronous output; the signal preprocessing module is used to perform amplitude processing on the first signal and the second signal; the protocol parsing module is used to parse the digital signal output by the sensor to be tested of the first signal, and send a first signal pulse to the timing module when the output indication signal of the sensor or the data header of the digital signal arrives; the level detection module is used to detect the second signal and send a second signal pulse to the timing module when the second signal jumps; the timing module is used to record the time of outputting the first signal pulse and the time of outputting the second signal pulse; the data recording and delay calculation module is used to calculate the difference between the time of outputting the first signal pulse and the time of outputting the second signal pulse, and obtain the output delay of the sensor to be tested. The present invention can complete the output delay calibration of different sensors.

Description

Technical Field

[0001] The present invention relates to the technical field of sensor delay calibration, and in particular to a sensor output delay calibration device and method. Background Art

[0002] In the fields of intelligent robots, autonomous navigation, and autonomous driving, due to the complexity and diversity of application scenarios, a single sensor solution can no longer meet the needs of applications, and multi-sensor fusion technology has been increasingly studied and applied in these fields. An important prerequisite for multi-sensor fusion is the time synchronization between different sensors. However, due to the differences in the working principles, electrical characteristics, operating frequencies, and digital signal preprocessing methods of different sensors, the delay of the physical signal from being detected to the output of the sensor is different. Therefore, developing a sensor delay calibration device and obtaining its output delay parameters are important for achieving multi-sensor data synchronization, improving the accuracy of multi-modal information fusion, and promoting the widespread application of intelligent robots.

[0003] In the application of multimodal information fusion, the traditional methods of synchronization or time alignment between sensors are: 1) roughly replacing the sampling time of the signal with the time when the processor receives the sensor data. 2) using software algorithms such as the cross-correlation method and the closest point iteration (ICP) method to determine the relative delay of different sensor data, but this method uses the time when the processor receives the sensor data as the calculation basis, and the reception time is often affected by factors such as the operating system scheduling, thereby adding additional uncertainty factors to the reception time, causing the results of the optimization algorithm to produce uncertainty deviations as the samples change. 3) using hardware trigger synchronization between similar sensors such as multi-channel cameras or multi-channel audio sensors. This method requires the hardware interface of the sensor to support the synchronization trigger signal provided by third-party hardware, or to support master-slave mode to achieve synchronization, so it is difficult to promote between non-similar sensors. 4) when the customized sensor outputs the sensor data, it also outputs the time information. Since there is no unified time reference between different sensors, it is necessary to add clock processing circuits to the processor periphery, which increases the complexity of the hardware system. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a sensor output delay calibration device and method, which can complete the output delay calibration of different sensors.

[0005] The technical solution adopted by the present invention to solve its technical problem is: to provide a sensor output delay calibration device, including a synchronization signal generating device, a signal preprocessing module, a level detection module, a protocol parsing module, a timing module and a data recording and delay calculation module; the synchronization signal generating device is used to generate a first signal and a second signal for synchronous output; the first signal is a sudden change detection signal supported by the sensor to be tested, and the second signal is a level jump signal that can ignore the detection time; the signal preprocessing module is used to perform amplitude processing on the first signal and the second signal, and transmit the processed first signal to the sensor to be tested, and transmit the processed second signal to the level detection module; the protocol parsing module is used to parse the output of the sensor to be tested. The digital signal output by the sensor to be tested is read and a first signal pulse is sent to the timing module when the data header of the digital signal arrives; the level detection module is used to send a second signal pulse to the timing module when the second signal jumps; the timing module is used to record the time when the protocol analysis module outputs the first signal pulse and the time when the level detection module outputs the second signal pulse; the data recording and delay calculation module is used to record the digital signal output by the sensor to be tested and the output time of the corresponding first signal pulse, record the level state of each level jump in the second signal and the output time of the corresponding second signal pulse, and calculate the difference between the time of outputting the first signal pulse and the time of outputting the second signal pulse to obtain the output delay of the sensor to be tested.

[0006] The technical solution adopted by the present invention to solve the technical problem is: to provide a calibration method using the above sensor output delay calibration device, comprising the following steps:

[0007] (1) Initializing the sensor to be tested and obtaining the sampling frequency of the sensor to be tested, so that the sensor to be tested always maintains a fixed sampling period T for sampling signals during the calibration process;

[0008] (2) According to the sampling frequency of the sensor to be tested, a set of sensor mutation detection signal sequences and level signal sequences with a number of M is constructed using a synchronization signal generating device, and the time offset of adjacent signals is NT±T / K, where N is the number of sampling cycles, K is the calibration multiple, and K≤M≤2K;

[0009] (3) Calibrate the sudden signal output delay of the sensor to be tested corresponding to each signal in the sudden signal sequence, and calculate the output delay parameter of the sensor to be tested according to the output delay.

[0010] The occurrence time of the next signal in the signal sequence in step (2) is based on the occurrence time of the previous signal, and is delayed by N sampling periods T before increasing or decreasing the offset T / K.

[0011] The time interval NT±T / K between two adjacent signals in the signal sequence far exceeds the output delay of the sensor to be tested.

[0012] The calibration method is characterized in that it also includes determining whether the calibration data is normal, specifically: determining whether the output delay of the sensor to be tested first decreases to a minimum value with the difference of T / K, then suddenly increases, and continues to decrease with the difference of T / K, and repeats this cycle. If not, it indicates that the data is abnormal.

[0013] When calculating the output delay parameter of the sensor to be tested based on the output delay, when the signal duration in the mutation detection signal sequence is ≥ the sampling period T, take the a minimum data tdr(K)~tdr(K-a+1) before the calibration data changes from small to large and the a maximum data tdr(1)~tdr(a) after the calibration data changes from small to large, and calculate the output delay parameter of the sensor to be tested based on ((tdr(1)+…+tdr(a))+(tdr(K-a+1)+…+tdr(K)-aT)) / 2a; when the signal duration in the mutation detection signal sequence is < the sampling period T, select the minimum value of the calibration result as the output delay parameter of the sensor to be tested; wherein tdr(i) represents the i-th output delay in a single sampling period.

[0014] Beneficial Effects

[0015] Due to the adoption of the above-mentioned technical solution, the present invention has the following advantages and positive effects compared with the prior art: the present invention can realize the output delay calibration of different sensors in a low-power, low-cost embedded platform and obtain the delay parameters of different sensors under different time parameter settings, thereby providing key parameters for multimodal sensor fusion and synchronization. The present invention is suitable for integrated sensors packaged at the chip level and separate sensors packaged on printed circuit boards, and has certain versatility; the principle of the present invention is simple, and specific design and detail optimization can be performed according to different sensors; the measurement accuracy of the present invention can be adjusted according to actual needs, and improving the K value can improve the calibration accuracy. The present invention can provide a low-power and low-cost implementation solution for the application of multimodal sensors in the fields of intelligent robots, autonomous driving, autonomous navigation, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a structural schematic diagram of an embodiment of the present invention;

[0017] Figure 2 is a schematic diagram of a first signal and a second signal generated by a synchronization signal generating device in an embodiment of the present invention;

[0018] Figure 3 is a schematic diagram of a signal sequence constructed in an embodiment of the present invention;

[0019] Figure 4 is a theoretical value distribution diagram of sampling results corresponding to all signal sequences in an embodiment of the present invention;

[0020] Figure 5 is a schematic diagram of the structure of Embodiment 1 of the present invention;

[0021] Figure 6 It is a schematic diagram of the signal sequence constructed in Example 2 of the present invention. DETAILED DESCRIPTION

[0022] The present invention will be further described below in conjunction with specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and are not intended to limit the scope of the present invention. In addition, it should be understood that after reading the content taught by the present invention, those skilled in the art can make various changes or modifications to the present invention, and these equivalent forms fall within the scope limited by the appended claims of the application equally.

[0023] The process of detecting physical signals through sensors is as follows: 1) the sensor converts the physical signal to be measured into a weak electrical signal, 2) the front-end circuit pre-processes and amplifies the electrical signal, 3) the analog-to-digital converter (ADC for short) completes signal sampling and digitization, 4) digital signal processing (optional), 5) the digital interface transmits the digitized sensor data to the processor peripheral interface. Usually, before the peripheral interface circuit receives the sensor data and sends an interrupt request signal, the processor is unknown to the first four steps of the sensor data. The present invention is to calibrate the sum of the time of the first four steps of the sensor's working process, and call it the output time delay t of the sensor. dt .

[0024] The basic principle of the sensor output delay calibration method of the present invention is to construct two synchronous signal generating devices and two signal detection devices, and at the same time construct two signal detection paths to detect the two synchronous signals, path 1 is the path through the sensor to be tested, and path 2 is a reference path with negligible output delay. The time difference between the signals received by the detection device from the two signal detection paths is the output delay of the sensor to be tested.

[0025] In the two signal detection paths, the detected signals can be the same or different, but their signal sources must occur simultaneously. A typical example is that when lightning occurs, light signals and sound signals are output simultaneously, and the propagation delay of sound is estimated by ignoring the propagation time of light. The synchronous signal generating device for different sensors in the present invention is also based on this principle. The signal generating device simultaneously outputs at least two signals, one of which is a detection signal supported by the sensor, and the other is an electrical signal whose detection time can be ignored.

[0026] Among the two signals sent by the synchronization signal generating device, the sensor signal needs to have a mutation feature to facilitate the corresponding signal feature in the subsequent delay calculation; the other signal is the simplest level jump signal, and its level jump moment is used to represent the corresponding occurrence time of the sensor mutation feature signal.

[0027] The embodiment of the present invention relates to a sensor output delay calibration device, such as Figure 1 As shown, this embodiment aims at the time synchronization requirement of different sensor data in multimodal sensor fusion, and designs a sensor output delay calibration device (referred to as delay calibration device) including a synchronization signal generating device, a signal preprocessing module, a level detection module, a protocol parsing module, a timing module and a data recording and delay calculation module.

[0028] The synchronization signal generating device is used to generate a first signal and a second signal for synchronous output; the first signal is a mutation detection signal supported by the sensor to be tested, referred to as a sensor mutation signal, and the second signal is a level jump signal that can ignore the detection time, referred to as a level signal. The two signals can use the same signal in special circumstances.

[0029] The signal preprocessing module is used to perform amplitude processing on the first signal and the second signal, and transmit the processed first signal to the sensor to be tested, and transmit the processed second signal to the level detection module. The processed first signal is adapted to the input range of the sensor to be tested, and the processed second signal is adapted to the level standard of the level detection module.

[0030] The protocol analysis module is used to analyze the digital signal output by the sensor to be tested, and send a first signal pulse to the timing module when the output indication signal of the sensor or the data header of the digital signal arrives.

[0031] The level detection module is used to send a second signal pulse to the timing module when the second signal jumps. It can be implemented based on a comparator, and its detection delay is less than the circuit working clock cycle.

[0032] The timing module is used to record the time when the protocol analysis module outputs the first signal pulse and the time when the level detection module outputs the second signal pulse.

[0033] The data recording and delay calculation module is used to record the digital signal output by the sensor to be tested and the output time of the corresponding first signal pulse, record the level state of each level jump in the second signal and the output time of the corresponding second signal pulse, and calculate the difference between the time of outputting the first signal pulse and the time of outputting the second signal pulse to obtain the output delay of the sensor to be tested.

[0034] Figure 2The situation where the synchronous signal generating device outputs the level jump signal and the sensor mutation characteristic signal at the same time is described. The judgment threshold of the mutation signal is defined as v th When the output value of the sensor exceeds v th When , the sampling signal is judged as a mutation signal. At the same time, the time when the mutation characteristic signal of the sensor exceeds the threshold is called the mutation signal duration t last Since the sampling time of the sensor is discrete, when the time when the synchronous signal generating device outputs the characteristic signal deviates from the sampling time of the ADC, the corresponding calibration result t dr Will include calibration error t de Therefore, a corresponding calibration method is needed to eliminate and reduce this error.

[0035] If the sensor mutation signal remains above the threshold value v th The high amplitude of the calibration signal is not obtained until the sampling time arrives, then the calibration mutation signal can be sampled, but the calibration result has a calibration error t de When the sensor's sudden change signal is a continuous shorter pulse signal or the high amplitude state does not last until the ADC sampling moment, the ADC cannot sample the sudden change signal, resulting in invalid calibration data.

[0036] Based on the above analysis, the sensor output delay calibration device of this embodiment includes the following three steps when performing output delay calibration:

[0037] 1) Initialize the sensor to be calibrated and obtain its sampling frequency so that the sensor to be tested always maintains a fixed period T sampling signal during the calibration process;

[0038] 2) If Figure 3 As shown, according to the sampling frequency of the sensor to be tested, a synchronous signal generating device is used to construct a set of signal sequences [s t1 ,…,s tm ,…,s tM ], where (1) the subscript tm indicates the time when the sensor mutation signal occurs; (2) t(m+1) = tm + N × T ± T / K, that is, the time when the next mutation signal occurs is based on the time when the previous mutation signal occurs, and is delayed by N sampling cycles T before adding or subtracting the offset T / K. It is recommended that K be a positive integer greater than 10. (3) t(m+1) - tm = N × T ± T / K >> t dt, that is, the time interval between two adjacent sensor mutation signals in the signal sequence is long enough and far exceeds the sensor output delay. At this time, the first sensor mutation signal output after the mutation signal can be used for sensor output delay calibration; (4) The number of signal sequences M>K, satisfying T / K×M>T; (5) Under the premise of being able to clearly determine the mutation signal, the threshold should be adjusted to obtain the longest possible duration of the mutation signal. It is better that the duration of the mutation signal is greater than the sampling period T. When the duration of the mutation signal is less than T / K, it may cause all calibrations to be invalid.

[0039] 3) Calibrate the sensor output delay corresponding to each signal in the signal sequence, that is, the delay corresponding to the first sensor output mutation data after the sensor mutation signal occurs. Based on the output delay calibration results of each signal, the output delay parameters of the sensor are calculated. The theoretical value distribution of the sampling results corresponding to all signals is as follows: Figure 4 As shown, it is a periodic discrete sequence in general, and a decreasing arithmetic sequence within a single period. dr-(i) Mark the sequence in a single period, the number of sequences is M, where t dr-(1) is the maximum value, t dr-(M) is the minimum value.

[0040] When the duration of the sudden change signal is ≥ the sampling period T, the theoretical value of the difference between the maximum and minimum values ​​of the calibration result is (K–1)T / K. Each single calibration can output a valid calibration result, corresponding to Figure 4 All solid points in the table. The correct calibration results should have the following characteristics: (1) First, the difference of T / K decreases to the minimum value, then suddenly increases to the maximum value, and continues to decrease with the difference of T / K, and so on; (2) After the calibration results in the next sampling cycle are subtracted from one sampling cycle, as shown by the hollow points, the sampling results in the previous and next cycles form an arithmetic progression of M in the calibration order and decrease with the difference of T / K. If the sampling data does not meet the above two points, it can be judged as abnormal data.

[0041] Select the smallest a and the largest a calibration results from all calibration results, and obtain the calibration results using the following formula:

[0042] ((t dr(M-a+1) +…+t dr(M) )+(t dr(1) +…+t dr(a) –aT)) / 2a——Formula (1)

[0043] The output delay error in this calibration result is ±T / 2K.

[0044] When the duration of the sudden signal is less than the sampling period T (not recommended), it may cause Figure 4 The t at the large end of a single cycle dr(1)……t dr(a) Missing. The minimum value can be used as the output calibration result. When the minimum value is used as the output delay calibration result, the output delay error is ±T / K.

[0045] When the duration of the sudden signal is less than the sampling period T / K, it may be as follows Figure 4 The minimum value in is not output either. At this time, the calibration fails and it is necessary to further increase the K value and recalibrate.

[0046] When K ≥ 10, the calibration error is one order of magnitude smaller than the sensor sampling period, and the calibration accuracy can meet most applications.

[0047] The present invention is further illustrated by the following specific examples.

[0048] In specific applications, 1) the signal preprocessing module is implemented in the board-level circuit and adjusts the signal amplitude according to different situations; 2) the level detection module, protocol analysis module, and high-precision timing module can be implemented based on the field programmable gate array (FPGA for short). FPGA can also add PC communication interface and characteristic signal generation module according to actual needs. Among them, the protocol analysis module corresponds to the digital interface of the sensor, usually SPI, IIC, UART, etc.; the clock frequency of the timing module determines the calibration time accuracy, such as when the working clock is 100MHz, the timing accuracy is 10 nanoseconds; the PC communication interface module completes the correspondence between the two signal data and the time mark and communicates with the PC. 3) The data recording and delay calculation module is completed by the PC program.

[0049] like Figure 3 As shown, during the calibration process, the sensor is always kept sampling signals at a fixed period. When the interval of the signal generated by the synchronous signal generating device is an integer multiple of the sampling period (as shown by the dotted arrow in the attached figure), there is always a fixed calibration error between the calibration result and the true value; after the synchronous signal generating device generates the first signal, it continues to send M-1 identical signals, and the interval between each signal and the previous signal is T(N+1 / K) (K=10 in the attached figure), that is, the phase of the sensor mutation signal time in the sensor sampling period changes uniformly. When the number of signals generated by the synchronous signal generating device M is greater than K, the output delay calibration error of the minimum value in the calibration result is ±T / K.

[0050] Common sensors can be divided into two categories according to the degree of integration. One category packages the front-end physical signal-to-electrical signal circuit and the back-end ADC circuit together, while the other category separates the two. The former often uses chip-level packaging, such as the IMU chip based on MEMS technology, which is called an integrated sensor in this embodiment; the latter often uses board-level packaging, such as the MIC head and sound encoder connected by a circuit board, which is called a separate sensor in this embodiment.

[0051] Example 1: IMU acceleration axis output delay calibration device (see Figure 5 )

[0052] In the design of the output delay calibration device of the integrated sensor, the construction of the synchronous signal generating device is the key. In this example, a slight impact event is used to make the IMU generate a sensor mutation signal, and a metal switch contact is designed at the impact junction. When the delay detection device of the IMU to be tested is collided, the acceleration value of the IMU mutates, and the contact is turned on and inputs a level jump signal to the calibration device. In addition, a timing control device is added to the synchronous signal generating device to control the impact moment, so as to construct a signal sequence that meets the calibration time requirements.

[0053] The level detection module in the FPGA receives the jump signal of the metal contact input, and the trigger timing module captures the jump moment of the level signal. The jump moment signal will be transmitted to the PC through the PC communication module. On the other hand, the protocol analysis module receives data through the IIC interface or SPI interface of the IMU. Each set of data received sends a pulse through the hardware circuit, and the trigger timing module captures the corresponding receiving time of the data. Each set of acceleration data and the receiving time generated by the trigger timing module are transmitted to the PC through the PC communication module. Finally, the delay result is calculated by the delay calculation module on the PC side.

[0054] Here we expand the description of the relationship between the signal sequence and the sampling process. If the sampling frequency of the IMU is 1000Hz (i.e., the sampling period is 1ms), and K=10, the change in the time interval between two adjacent collision signals is 1ms / 10=0.1ms. Assuming that the IMU takes the first sampling after the first collision 0.75ms after the first collision, if the time of the second collision is controlled within N sampling periods, i.e., N ms, then the first sampling after the second collision still occurs 0.75ms after the second collision; if the time of the second collision is controlled within N sampling periods, i.e., N ms, and then moves forward 0.1ms, then the time between the second collision and the first sampling after the collision is shortened to 0.65ms. By constructing collision events in the same way, the time between the 7th collision and the first sampling after the collision is shortened to 0.15ms, and the 8th collision is further shortened to 0.05ms. However, at the 9th collision, the sampling point is just missed and only the next sampling can be used. The time from the collision to the most recent sampling becomes 0.95ms, and subsequent signals decrease in sequence.

[0055] The calibration data of different signals are divided based on the time of each impact. The calibration results of a single set of data can be obtained by comparing the time when the level signal jumps and the time when the IMU signal suddenly changes in each set of data.

[0056] After completing the calibration of all signals, the calibration results need to be analyzed and preprocessed. When the difference between the maximum and minimum values ​​of the calibration results is approximately (K–1)T / K, it indicates that the duration of the mutation signal is ≥ the sampling period T, and the sensor output delay parameter can be calculated using formula (1). Otherwise, the minimum calibration result is selected as the sensor output delay calibration parameter.

[0057] Example 2: Calibration of audio codec output delay

[0058] If you only care about the output delay generated by the sensor's ADC and digital signal processing circuit, you can use a separate sensor output delay calibration device and replace the synchronization signal generation device with a characteristic signal generation module, which is easier to implement. Figure 5 As shown. (1) The characteristic signal generation module is used to generate a non-periodic square wave signal. Here, the jump signal of the square wave signal is used as the sensor mutation signal of the Codec. At this time, the sensor mutation signal and the level signal are the same. (2) The signal preprocessing module performs amplitude processing on the square wave signal to adapt it to the input range of the sensor to be tested. Since the sensor signal is weak, the signal preprocessing module can use a resistor divider circuit.

[0059] Similar to the output delay calibration of integrated sensors, the output delay calibration process of separate sensors also requires the construction of a set of signals with increasing phases during the sampling period.

[0060] The construction method of pulse signal sequence is as follows Figure 6 As shown in the figure: (1) The pulse width is fixed and greater than a sampling period T. After the rising edge of the pulse signal arrives, the first sampling time can definitely sample a high level, and the sampling data before the pulse signal arrives must be a low level; (2) The interval between the rising edges of two adjacent pulse signals should be much larger than the sensor delay. The interval between the latter signal and the previous signal is T×N+T clk , where T clk It represents a single clock cycle, that is, the increment of the signal interval is a single clock cycle, and the increment range increases from a single clock cycle to twice the clock cycle; (3) If the working clock is 100Mhz, the theoretical calibration accuracy can reach ±5ns.

Claims

1. A sensor output delay calibration device, characterized in that: It includes a synchronization signal generating device, a signal preprocessing module, a level detection module, a protocol analysis module, a timing module and a data recording and delay calculation module; the synchronization signal generating device is used to generate a first signal and a second signal for synchronous output; the first signal is a sudden change detection signal supported by the sensor to be tested, and the second signal is a level jump signal that can ignore the detection time; The signal preprocessing module is used to perform amplitude processing on the first signal and the second signal, and transmit the processed first signal to the sensor to be tested, and transmit the processed second signal to the level detection module; the protocol analysis module is used to analyze the digital signal output by the sensor to be tested, and send a first signal pulse to the timing module when the data header of the digital signal arrives; the level detection module is used to send a second signal pulse to the timing module when the second signal jumps; the timing module is used to record the time when the protocol analysis module outputs the first signal pulse and the time when the level detection module outputs the second signal pulse; the data recording and delay calculation module is used to record the digital signal output by the sensor to be tested and the output time of the corresponding first signal pulse, record the level state of each level jump in the second signal and the output time of the corresponding second signal pulse, and calculate the difference between the time of outputting the first signal pulse and the time of outputting the second signal pulse to obtain the output delay of the sensor to be tested.

2. A calibration method using the sensor output delay calibration device as claimed in claim 1, characterized in that: The following steps are involved: (1) Initializing the sensor to be tested and obtaining the sampling frequency of the sensor to be tested, so that the sensor to be tested always maintains a fixed sampling period T for sampling signals during the calibration process; (2) According to the sampling frequency of the sensor to be tested, a set of sensor mutation detection signal sequences and level signal sequences with a number of M is constructed using a synchronization signal generating device, and the time offset of adjacent signals is NT±T / K, where N is the number of sampling cycles, K is the calibration multiple, and K≤M≤2K; (3) Calibrate the sudden signal output delay of the sensor to be tested corresponding to each signal in the sudden signal sequence, and calculate the output delay parameter of the sensor to be tested according to the output delay.

3. The calibration method according to claim 2, characterized in that: The occurrence time of the next signal in the signal sequence in step (2) is based on the occurrence time of the previous signal, and is delayed by N sampling periods T before increasing or decreasing the offset T / K.

4. The calibration method according to claim 2, characterized in that: The time interval NT±T / K between two adjacent signals in the signal sequence far exceeds the output delay of the sensor to be tested.

5. The calibration method according to claim 2, characterized in that: It also includes determining whether the calibration data is normal, specifically: determining whether the output delay of the sensor to be tested first decreases to a minimum value with the difference of T / K, then suddenly increases, and continues to decrease with the difference of T / K, and repeats this cycle. If not, it indicates that the data is abnormal.

6. The calibration method according to claim 2, characterized in that: When calculating the output delay parameter of the sensor to be tested based on the output delay, when the signal duration in the mutation detection signal sequence is ≥ the sampling period T, take the a minimum data tdr(K)~tdr(K-a+1) before the calibration data changes from small to large and the a maximum data tdr(1)~tdr(a) after the calibration data changes from small to large, and calculate the output delay parameter of the sensor to be tested based on ((tdr(1)+…+tdr(a))+(tdr(K-a+1)+…+tdr(K)-aT)) / 2a; when the signal duration in the mutation detection signal sequence is < the sampling period T, select the minimum value of the calibration result as the output delay parameter of the sensor to be tested; wherein tdr(i) represents the i-th output delay in a single sampling period.

Images