Ship lock water level monitoring system and method based on cascade arrangement scheme

Abstract

The invention discloses a ship lock water level monitoring system and method based on a cascade arrangement scheme. The system comprises a cascade water level measuring structure composed of a plurality of fiber grating sensors. The plurality of fiber grating sensors are connected to a demodulator through a fiber fusion box, and an upper computer is integrated in the demodulator; the demodulator is used for controlling laser emission and grating reflected light receiving of the fiber grating sensor and demodulating collected data to obtain demodulated data; the upper computer is used for performing data processing on the demodulated data, converting the demodulated data into a water level value, interacting with an operator and communicating with the central control system, the water level gauge is verified and calibrated according to the characteristics of a cascade arrangement scheme, and the verified effective data is fused and filtered through a combined Kalman filtering method. The ship lock water level change characteristics are effectively combined, the advantages of a fiber grating measurement system are utilized, and the accuracy and reliability of ship lock water level measurement are achieved.

Description

Technical Field

[0001] The invention relates to the field of ship lock water level measurement, and in particular to a ship lock water level monitoring system and method based on a cascade arrangement scheme. Background Art

[0002] In the industrial control system of the ship lock, the ship lock operation water level monitoring system plays a vital role. As the core unit of the ship lock operation water level monitoring system, the reliability and accuracy of the water level measurement value output by the water level measurement system, referred to as the "water level signal", directly affects whether the ship lock equipment can operate safely and smoothly and whether the ship can smoothly shift and berth, and is one of the key factors determining the safety of the ship lock operation; the water level signal is not only used to judge important operating parameters such as the ship lock operation mode, but the water level signal difference also participates in the key process control of the ship lock operation such as the flat pressure opening and closing of the ship lock gate, water level protection, valve acoustic vibration suppression, and reverse head suppression.

[0003] Since its emergence, fiber grating sensing technology has been increasingly valued by the scientific and engineering communities for its unique advantages such as natural resistance to electromagnetic interference, corrosion resistance, and easy connection and arrangement. It has been widely used in information and sensing systems, such as liquid position detection of pipette needles in the medical industry; monitoring of tail flame temperature during rocket launches or fighter takeoffs; measurement of corresponding strain distances in flutter detection of airborne drooping tail models in wind tunnels; real-time monitoring of surrounding pressure on satellite rockets and other surfaces; real-time monitoring of structural deformation during deep diving; and the need to obtain acceleration parameters when aircraft carriers take off, all of which demonstrate its reliability and accuracy under extreme conditions. In these harsh environments with strong electromagnetic interference, it is particularly convenient to use passive devices such as optical fibers to measure parameters. Therefore, fiber grating technology liquid level sensors have gradually attracted the interest of researchers in the water conservancy industry. In recent years, researchers have conducted some useful explorations on the application of fiber grating in the field of lock water level measurement.

[0004] However, the research direction of domestic and foreign researchers on fiber grating liquid level sensors mainly focuses on the performance of the fiber grating sensor itself. There are few related studies on fiber grating water level measurement systems in the use scenarios of ship locks, especially high-head ship locks. In actual application, traditional means are still used. The measurement accuracy and reliability are not outstanding in scenarios where the water level fluctuation in the ship lock is large, the water level change rate is high, and the measurement environment is harsh. Therefore, it is necessary to design a ship lock water level monitoring system and method based on a cascade arrangement scheme to solve the above problems. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a ship lock water level monitoring system and method based on a cascade arrangement scheme, which effectively combines the ship lock water level change characteristics and utilizes the advantages of a fiber grating measurement system to achieve accuracy and reliability in ship lock water level measurement.

[0006] In order to achieve the above technical effects, the technical solution adopted by the present invention is:

[0007] A ship lock water level monitoring system based on a cascade arrangement scheme includes a plurality of fiber grating sensors, wherein the plurality of fiber grating sensors are respectively connected to the inner wall of a water level measurement well through independent mounting brackets; the plurality of fiber grating sensors are electrically connected through optical fibers, and the optical fibers are connected to a demodulator through an optical fiber fusion box, and a host computer is integrated in the demodulator;

[0008] The fiber fusion box is used to perform fiber fusion splicing when the fiber is connected to the demodulator;

[0009] The demodulator is used to control the laser emission of the fiber grating sensor, the reception of the grating reflected light, and to demodulate the collected data to obtain the demodulated data;

[0010] The host computer is used to process the demodulated data, convert it into water level values, interact with the operator, and communicate with the central control system.

[0011] Preferably, the above method of a ship lock water level monitoring system based on a cascade arrangement scheme comprises:

[0012] S1, the host computer performs data preprocessing on the data demodulated by the demodulator through limiting filtering, removes gross errors in the data, and converts the demodulated data into water level values;

[0013] S2, the water level measurement well is filled or drained. When a fiber Bragg grating sensor enters or leaves the water body, the water level value fed back by the current fiber Bragg grating sensor is verified and calibrated using the installation elevation of the current fiber Bragg grating sensor, and the zero point and slope of the fiber Bragg grating sensor are calibrated;

[0014] S3, using the joint Kalman filter algorithm to perform data fusion on the verified water level value data;

[0015] S4, outputs the fused water level value.

[0016] Preferably, in step S2, the water level value fed back by the current fiber Bragg grating sensor is verified and calibrated using the installation elevation of the current fiber Bragg grating sensor. The method for calibrating the zero point and slope of the fiber Bragg grating sensor is:

[0017] S201, when water level measurement is performed for filling or draining, the water level changes upward or downward, and a fiber grating sensor S is detected. i entering or leaving a body of water;

[0018] S202, using the fiber Bragg grating sensor S i The installation height H is taken as the standard water level value to calculate the fiber Bragg grating sensor S iFiber Bragg grating sensor S below the installation position i-1 The difference between the value of and the standard water level value H, that is, calculate |S i-1 -H|;

[0019] S203, judgment | S i-1 -H|Is it greater than the set allowable error D? If it is greater than the allowable error, the fiber Bragg grating sensor S i-1 The value is marked as an invalid water level, and the standard water level value H is used as the calibration value to adjust the fiber Bragg grating sensor S i-1 The zero point of |S i-1 -H| can fall within the allowable error D; if it is less than the allowable error D, the fiber Bragg grating sensor S i-1 The value of is marked as the effective water level;

[0020] S204, repeating steps S202 and S203 until the fiber Bragg grating sensor S i All the following fiber grating sensors have completed the judgment and calibration process;

[0021] S205, repeating steps S201 to S204 until the water filling or draining process is completed.

[0022] Preferably, in step S205, if the number of zero point calibrations of a fiber Bragg grating sensor reaches two times, a slope calibration is performed on the water level, and the slope calibration method is:

[0023]

[0024] Where ΔH2 is the zero point calibration value during the second zero point calibration, H1 and H2 are the water level values ​​during the first and second calibrations, respectively. 旧 is the original slope.

[0025] Furthermore, in the loop of step S205, if the number of zero point calibrations of a fiber Bragg grating sensor reaches three or more times, it is considered that the fiber Bragg grating sensor has a hardware failure that cannot be solved by calibration, and the host computer sends a fault warning signal to notify the staff to handle it.

[0026] Preferably, in step S3, the method of using the joint Kalman filter algorithm to fuse the verified water level value data is:

[0027] S301, perform state prediction on the data marked as effective water level and calculate the prior prediction value X of the water level i -(k):

[0028]

[0029] The discrete dynamic equation of water level change is:

[0030] X(k)=AX(k-1)+W(k);

[0031] The observation equation is:

[0032] Z i (k)=HX(k)+V i (k);

[0033] A is the transfer matrix in the dynamics equation, is the posterior estimate of the water level at the previous moment, represents the prior prediction value of the water level value; X(k) is the state variable in the water level change state equation, W(k) is the system process noise, and Z i (k) represents the observation vector of the i-th water level gauge, H represents the observation matrix, V i (k) represents the measurement noise, k represents the sampling sequence;

[0034] S302, calculate the prior prediction value of the covariance:

[0035]

[0036] in represents the covariance matrix of the previous moment, represents the prior prediction of the covariance matrix, Q i (k) is the noise variance of the system, A T is the transposed matrix of the transfer matrix in the dynamic equation;

[0037] S303, calculate the Kalman gain value:

[0038]

[0039] Where R i (k) represents the variance of the fiber Bragg grating sensor measurement noise; G i (k) represents the Kalman gain value;

[0040] S304, update the covariance matrix of the state variables to prepare for the next cycle:

[0041]

[0042] in Represents the covariance matrix at the current moment;

[0043] S305, update the posterior estimate of the water level value:

[0044]

[0045] in represents the posterior estimate of the water level, Represents the prior prediction value of the water level; Z i represents the observation vector of the i-th water level gauge;

[0046] S306, looping steps S301 to S305 in each sampling period to provide a filtered estimated value for each effective water level;

[0047] S307, combining all fiber Bragg grating sensor information to perform an overall estimation of the water level:

[0048]

[0049]

[0050] Among them, X + (k) represents the overall posterior estimate of the water level, P + (k) is the intermediate variable for calculating the distribution coefficient of each water level gauge, is the inverse of the covariance matrix of the i-th water level gauge at the current moment;

[0051] The distribution relationship of the system process information Q between each local filter is:

[0052]

[0053]

[0054] in, represents the inverse of the noise variance of the i-th water level gauge process; β i It represents the allocation coefficient, which reflects the weights of different fiber Bragg grating sensors. When the credibility of a fiber Bragg grating sensor decreases, the allocation coefficient decreases.

[0055] The beneficial effects of a ship lock water level monitoring system and method based on a cascade arrangement scheme provided by the present invention are as follows:

[0056] 1. The present invention is aimed at measuring the water level difference between inside and outside the ship lock chamber or other water level difference measurement scenarios. The cascade solution can increase the measurement accuracy in principle.

[0057] 2. The present invention can utilize all effective water level meter data by using the joint Kalman filter algorithm, thus avoiding data waste and improving the accuracy and stability of water level measurement values ​​from the algorithm perspective;

[0058] 3. The present invention utilizes the dynamic change of the joint Kalman filter allocation coefficient. When the water level meter data is abnormal, the allocation coefficient can be lowered, which can effectively avoid the final water level result being affected by the error of a certain water level meter data;

[0059] 4. The zero point calibration and slope calibration algorithms designed by the present invention using the layout characteristics of the cascade scheme can effectively calibrate the zero point drift and slope drift of the water level meter, ensuring the long-term measurement accuracy of the water level meter. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 The water level measurement system structure and data processing flow chart of the present invention;

[0061] Figure 2 It is a schematic diagram of the cascade arrangement of the fiber Bragg grating sensor of the present invention;

[0062] Figure 3 It is a schematic diagram of the zero point calibration algorithm flow of the present invention;

[0063] Figure 4 Schematic diagram of the slope calibration algorithm flow of the present invention. DETAILED DESCRIPTION

[0064] Embodiment 1:

[0065] like Figure 1 As shown, a ship lock water level monitoring system based on a cascade arrangement scheme includes a plurality of fiber grating sensors, and the plurality of fiber grating sensors are respectively connected to the inner wall of a water level measurement well through independent mounting brackets; the plurality of fiber grating sensors are electrically connected through optical fibers, and the optical fibers are connected to a demodulator through an optical fiber fusion box, and a host computer is integrated in the demodulator;

[0066] The fiber fusion box is used to perform fiber fusion splicing when the fiber is connected to the demodulator;

[0067] The demodulator is used to control the laser emission of the fiber grating sensor, the reception of the grating reflected light, and to demodulate the collected data to obtain the demodulated data;

[0068] The host computer is used to process the demodulated data, convert it into water level values, interact with the operator, and communicate with the central control system.

[0069] Furthermore, if Figure 2 As shown, fiber grating sensors with increasing ranges are arranged from top to bottom at a ship lock water level measurement point, the installation elevation of each fiber grating sensor is determined, and the upper limit of the allowable error is set according to the actual situation.

[0070] Embodiment 2:

[0071] The above-mentioned method of the ship lock water level monitoring system based on the cascade arrangement scheme comprises:

[0072] S1, the host computer performs data preprocessing on the data demodulated by the demodulator through limiting filtering to remove gross errors in the data and convert the demodulated data into water level values;

[0073] S2, the water level measurement well is filled or drained. When a fiber Bragg grating sensor enters or leaves the water body, the water level value fed back by the current fiber Bragg grating sensor is verified and calibrated using the installation elevation of the current fiber Bragg grating sensor, and the zero point and slope of the fiber Bragg grating sensor are calibrated;

[0074] S3, using the joint Kalman filter algorithm to perform data fusion on the verified water level value data;

[0075] S4, outputs the fused water level value.

[0076] like Figure 3 As shown, preferably, in step S2, the water level value fed back by the current fiber Bragg grating sensor is verified and calibrated using the installation elevation of the current fiber Bragg grating sensor, and the method for calibrating the zero point and slope of the fiber Bragg grating sensor is:

[0077] S201, when water level measurement is performed for filling or draining, the water level changes upward or downward, and a fiber grating sensor S is detected. i entering or leaving a body of water;

[0078] S202, using the fiber Bragg grating sensor S i The installation height H is taken as the standard water level value to calculate the fiber Bragg grating sensor S i Fiber Bragg grating sensor S below the installation position i-1 The difference between the value of and the standard water level value H, that is, calculate |S i-1 -H|;

[0079] S203, judgment | S i-1 -H|Is it greater than the set allowable error D? If it is greater than the allowable error, the fiber Bragg grating sensor S i-1 The value is marked as an invalid water level, and the standard water level value H is used as the calibration value to adjust the fiber Bragg grating sensor S i-1 The zero point of |S i-1 -H| can fall within the allowable error D; if it is less than the allowable error D, the fiber Bragg grating sensor S i-1 The value of is marked as the effective water level;

[0080] S204, repeating steps S202 and S203 until the fiber Bragg grating sensor S i All the following fiber grating sensors have completed the judgment and calibration process;

[0081] S205, repeating steps S201 to S204 until the water filling or draining process is completed.

[0082] like Figure 4As shown, preferably, in step S205, if the number of zero point calibrations of a fiber Bragg grating sensor reaches two times, the water level is subjected to slope calibration, and the slope calibration method is:

[0083]

[0084] Where ΔH2 is the zero point calibration value during the second zero point calibration, H1 and H2 are the water level values ​​during the first and second calibrations, respectively. 旧 is the original slope.

[0085] Furthermore, in the loop of step S205, if the number of zero point calibrations of a fiber Bragg grating sensor reaches three or more times, it is considered that the fiber Bragg grating sensor has a hardware failure that cannot be solved by calibration, and the host computer sends a fault warning signal to notify the staff to handle it.

[0086] Preferably, in step S3, the method of using the joint Kalman filter algorithm to fuse the verified water level value data is:

[0087] S301: Perform state prediction on the data marked as effective water level and calculate the prior prediction value of the water level.

[0088]

[0089] The discrete dynamic equation of water level change is:

[0090] X(k)=AX(k-1)+W(k);

[0091] The observation equation is:

[0092] Z i (k)=HX(k)+V i (k);

[0093] A is the transfer matrix in the dynamics equation, is the posterior estimate of the water level at the previous moment, represents the prior prediction value of the water level value; X(k) is the state variable in the water level change state equation, W(k) is the system process noise, and Z i (k) represents the observation vector of the i-th water level gauge, H represents the observation matrix, V i (k) represents the measurement noise, k represents the sampling sequence;

[0094] S302, calculate the prior prediction value of the covariance:

[0095]

[0096] in represents the covariance matrix of the previous moment, represents the prior prediction of the covariance matrix, Q i (k) is the noise variance of the system, A T is the transposed matrix of the transfer matrix in the dynamic equation;

[0097] S303, calculate the Kalman gain value:

[0098]

[0099] Where R i (k) represents the variance of the fiber Bragg grating sensor measurement noise; G i (k) represents the Kalman gain value;

[0100] S304, update the covariance matrix of the state variables to prepare for the next cycle:

[0101]

[0102] in Represents the covariance matrix at the current moment;

[0103] S305, update the posterior estimate of the water level value:

[0104]

[0105] in represents the posterior estimate of the water level, Represents the prior prediction value of the water level; Z i represents the observation vector of the i-th water level gauge;

[0106] S306, looping steps S301 to S305 in each sampling period to provide a filtered estimated value for each effective water level;

[0107] S307, combining all fiber Bragg grating sensor information to perform an overall estimation of the water level:

[0108]

[0109]

[0110] Among them, X + (k) represents the overall posterior estimate of the water level, P + (k) is the intermediate variable for calculating the distribution coefficient of each water level gauge, is the inverse of the covariance matrix of the i-th water level gauge at the current moment;

[0111] The distribution relationship of the system process information Q between each local filter is:

[0112]

[0113]

[0114] in, represents the inverse of the noise variance of the i-th water level gauge process; β i It represents the allocation coefficient, which reflects the weights of different fiber Bragg grating sensors. When the credibility of a fiber Bragg grating sensor decreases, the allocation coefficient decreases.

[0115] Embodiment three:

[0116] Taking the zero point calibration and slope calibration during a single ship lock filling process from the lowest water level to the highest water level in a lock chamber as an example, the algorithm flow and calibration method are explained.

[0117] 1. The ship lock is in the water-filled state. At this time, the value of S6 fiber grating sensor is valid, and the other fiber grating sensors are not in the water.

[0118] 2. The water level changes upward. At this time, the water level reaches the position of the S5 fiber Bragg grating sensor. The S5 data shows the signal characteristics of entering the water body. The actual water level value is judged to be H5. The size relationship between |S6-H5| and D is judged. If it is less than the allowable error D, it is determined to be a valid fiber Bragg grating sensor. If the error is greater than D, the zero point of S6 is adjusted according to the current water level value of S5, which is H5.

[0119] 3. The water level continues to change upward and reaches the position of the S4 fiber Bragg grating sensor. The S4 data shows the signal characteristics of entering the water body. The actual water level value is determined to be H4. The size relationship between |S6-H4|, |S5-H4| and D is determined. If it is less than the allowable error D, it is determined to be a valid fiber Bragg grating sensor. If the error is greater than D, the zero point of the fiber Bragg grating sensor is adjusted according to the current water level value H4.

[0120] 4. The water level continues to change upward and reaches the position of the fiber Bragg grating sensor S3. The S3 data shows the signal characteristics of entering the water body. The actual water level value is determined to be H3. The size relationship between |S6-H3|, |S5-H3|, |S4-H3|, and D is determined. If it is less than the allowable error D, it is determined to be a valid fiber Bragg grating sensor. If the error is greater than D, the zero point of the fiber Bragg grating sensor is adjusted according to the current water level value H5.

[0121] 5. Repeat this process until the water level exceeds H2, which means the calibration is finished.

[0122] In the above example of zero point calibration process, S6 will judge 4 times, S5 will judge 3 times, S4 will judge 2 times, S3 will judge 1 time, and S2 will judge 0 times; if the number of calibrations of a fiber Bragg grating sensor reaches 2 times, that is, according to Figure 4 The slope correction method is used to adjust the slope. If a fiber Bragg grating sensor is calibrated more than twice, that is, the fiber Bragg grating sensor is still judged to need calibration after zero point calibration and slope calibration, it is considered that the fiber Bragg grating sensor is faulty; the calibration process during the water discharge process is the same as the water filling process mentioned above.

[0123] In the above zero point correction and slope correction algorithms, the lower the position of the fiber Bragg grating sensor is and the larger the range is, the more correction judgments are required. This is in line with the experience that sensors in large range and deep water environments are more likely to drift, and can effectively ensure the data reliability of fiber Bragg grating sensors in harsh environments.

Claims

1. A ship lock water level monitoring system based on a cascade arrangement scheme, characterized in that: It includes multiple fiber grating sensors, which are respectively connected to the inner wall of the water level measurement well through independent mounting brackets; multiple fiber grating sensors are electrically connected through optical fibers, and the optical fibers are connected to a demodulator through an optical fiber fusion box, and a host computer is integrated in the demodulator; The fiber fusion box is used to perform fiber fusion splicing when the fiber is connected to the demodulator; The demodulator is used to control the laser emission of the fiber grating sensor, the reception of the grating reflected light, and to demodulate the collected data to obtain the demodulated data; The host computer is used to process the demodulated data, convert it into water level values, interact with the operator, and communicate with the central control system.

2. A method for monitoring the water level of a ship lock based on a cascade arrangement scheme according to claim 1, characterized in that: include: S1, the host computer performs data preprocessing on the data demodulated by the demodulator through limiting filtering, removes gross errors in the data, and converts the demodulated data into water level values; S2, the water level measurement well is filled or drained. When a fiber Bragg grating sensor enters or leaves the water body, the water level value fed back by the current fiber Bragg grating sensor is verified and calibrated using the installation elevation of the current fiber Bragg grating sensor, and the zero point and slope of the fiber Bragg grating sensor are calibrated; S3, using the joint Kalman filter algorithm to perform data fusion on the verified water level value data; S4, outputs the fused water level value.

3. The method of a ship lock water level monitoring system based on a cascade arrangement scheme according to claim 2, characterized in that: In step S2, the water level value fed back by the current fiber Bragg grating sensor is verified and calibrated using the installation elevation of the current fiber Bragg grating sensor. The method for calibrating the zero point and slope of the fiber Bragg grating sensor is: S201, when water level measurement is performed for filling or draining, the water level changes upward or downward, and a fiber grating sensor S is detected. i entering or leaving a body of water; S202, using the fiber Bragg grating sensor S i The installation height H is taken as the standard water level value to calculate the fiber Bragg grating sensor S i Fiber Bragg grating sensor S below the installation position i-1 The difference between the value of and the standard water level value H, that is, calculate |S i-1 -H|; S203, judgment | S i-1 -H|Is it greater than the set allowable error D? If it is greater than the allowable error, the fiber Bragg grating sensor S i-1 The value is marked as an invalid water level, and the standard water level value H is used as the calibration value to adjust the fiber Bragg grating sensor S i-1 The zero point of |S i-1 -H| can fall within the allowable error D; if it is less than the allowable error D, the fiber Bragg grating sensor S i-1 The value of is marked as the effective water level; S204, repeating steps S202 and S203 until the fiber Bragg grating sensor S i All the following fiber grating sensors have completed the judgment and calibration process; S205, repeating steps S201 to S204 until the water filling or draining process is completed.

4. The method of the ship lock water level monitoring system based on the cascade arrangement scheme according to claim 3 is characterized in that: In step S205, if the number of zero point calibrations of a fiber Bragg grating sensor reaches two times, the water level is subjected to slope calibration, and the slope calibration method is as follows: Where ΔH2 is the zero point calibration value during the second zero point calibration, H1 and H2 are the water level values ​​during the first and second calibrations, respectively. 旧 is the original slope.

5. The method of the ship lock water level monitoring system based on the cascade arrangement scheme according to claim 3 is characterized in that: In step S3, the joint Kalman filter algorithm is used to perform data fusion on the verified water level data as follows: S301, perform state prediction on the data marked as effective water level and calculate the prior prediction value X of the water level i -(k): The discrete dynamic equation of water level change is: X(k)=AX(k-1)+W(k); The observation equation is: Z i (k)=HX(k)+V i (k); A is the transfer matrix in the dynamics equation, is the posterior estimate of the water level at the previous moment, represents the prior prediction of the kth water level value measured by the i-th water level gauge; X(k) is the state variable in the water level change state equation, W(k) is the system process noise, and Z i (k) represents the observation vector of the i-th water level gauge, H represents the observation matrix, V i (k) represents the measurement noise, k represents the sampling sequence; S302, calculate the prior prediction value of the covariance: in represents the covariance matrix of the previous moment, represents the prior prediction of the covariance matrix, Q i (k) is the noise variance of the system, A T is the transposed matrix of the transfer matrix in the dynamic equation; S303, calculate the Kalman gain value: Where R i (k) represents the variance of the fiber Bragg grating sensor measurement noise; G i (k) represents the Kalman gain value; S304, update the covariance matrix of the state variables to prepare for the next cycle: in Represents the covariance matrix at the current moment; S305, update the posterior estimate of the water level value: in represents the posterior estimate of the water level, represents the a priori predicted value of the water level; S306, looping steps S301 to S305 in each sampling period to provide a filtered estimated value for each effective water level; S307, combining all fiber Bragg grating sensor information to perform an overall estimation of the water level: Among them, X + (k) represents the overall posterior estimate of the water level, P + (k) is the intermediate variable for calculating the distribution coefficient of each water level gauge, is the inverse of the covariance matrix at the current moment; The distribution relationship of the system process information Q between each local filter is: in, represents the inverse of the noise variance of the i-th water level gauge process; β i It represents the allocation coefficient, which reflects the weights of different fiber Bragg grating sensors. When the credibility of a fiber Bragg grating sensor decreases, the allocation coefficient decreases.

Images