Ultrasonic imaging method, computer-readable medium, and ultrasonic detector

Abstract

The present invention relates to an ultrasonic imaging method, a computer-readable medium, and an ultrasonic detector. The relevant data obtained from a single transmission and reception process is input into an ultrasonic simulation model, and the obtained propagation time t i (x, z) and propagation time t j (x, z), combined with an imaging algorithm, to obtain an image of the detection surface of a second container filled with an inhomogeneous fluid; the ultrasonic simulation model is obtained as follows: a two-dimensional sound field is used to scan a first container filled with an inhomogeneous fluid, the first sound pressure in each transmission and reception process is obtained, an initial sound velocity field model is established, the second sound pressure is obtained by simulating the aforementioned transmission and reception process, and the gradient optimization algorithm is used for optimization, the second sound pressure - the first sound pressure

Description

Technical Field

[0001] The present invention relates to the technical field of ultrasonic detection of a container filled with non-uniform fluid, and particularly relates to an ultrasonic imaging method for a container filled with non-uniform fluid, a computer-readable medium storing an ultrasonic imaging program, and an ultrasonic detector. Background Art

[0002] For key connection parts such as the outgoing line device of an ultra-high voltage transformer, only relying on the natural circulation of insulating oil due to thermal expansion and contraction, it is in a semi-"dead oil area" state. Once a discharge breakdown occurs in this area, since the short-circuit current does not pass through the winding and the discharge energy is large, it becomes a high-risk area, seriously affecting the safety of the ultra-high voltage power grid.

[0003] Currently, the monitoring methods for key connection parts such as the outgoing line device of an ultra-high voltage transformer are divided into oil chromatogram detection route, iron core grounding current detection route, iron core grounding high-frequency partial discharge detection route, etc. These detection methods mainly detect the characteristics of the fault state and cannot effectively monitor the whole process of the internal state change of the transformer.

[0004] Theoretically, ultrasonic waves can be used for non-contact detection of the outgoing line device of an ultra-high voltage transformer. However, the insulating oil contained in the outgoing line device of the ultra-high voltage transformer will generate a thermal cycle due to thermal expansion and contraction during operation, resulting in changes in the acoustic characteristics of the insulating oil at different positions and different times, thus affecting the detection accuracy. Taking the single-channel ultrasonic vertical incidence scanning detection as an example, its detection time is too long, and the error caused by the acoustic characteristics of the insulating oil accumulates over time, resulting in extremely low accuracy. Taking the array ultrasonic scanning detection as an example, the array ultrasonic waves need to synthesize the sound beam, and it is difficult to achieve effective focusing, with low detection resolution and poor accuracy. Summary of the Invention

[0005] The purpose of the present invention is to provide an ultrasonic imaging method for a container filled with non-uniform fluid, a computer-readable medium storing an ultrasonic imaging program, and an ultrasonic detector to improve the accuracy of ultrasonic imaging of a container filled with non-uniform fluid.

[0006] The technical solution of the present invention is as follows:

[0007] An ultrasonic imaging method for a container filled with non-uniform fluid, comprising:

[0008] Step S20: Obtain the third emission position, n third sound pressure measurement positions during m emission and reception processes in the detection plane of a second container filled with non-uniform fluid, and the third sound pressure at the third sound pressure measurement position at the time t after emission, where m≥1 and n≥2;

[0009] Step S21: Calculate the propagation time of the ultrasonic wave emitted from the i-th third emission position during the m emission-reception processes to point P and the propagation time of the ultrasonic wave from point P to the j-th third sound pressure measurement position by using the ultrasonic simulation model corresponding to the detection plane of the second container filled with inhomogeneous fluid and the data obtained in step S21, and calculate the pixel value of point P by using the full focus imaging algorithm; traverse the points in the ultrasonic simulation model as point P to obtain the pixel values of all points in the ultrasonic simulation model, where i = 1, 2, …, m; j = 1, 2, …, n;

[0010] Step S22: Output the pixel values of all points in the ultrasonic simulation model;

[0011] The construction method of the ultrasonic simulation model includes the following steps:

[0012] Step S10: Scan the first container filled with inhomogeneous fluid in a circular motion within the detection plane. During each emission-reception process of scanning the first container filled with inhomogeneous fluid in a circular motion, emit the ultrasonic wave at the first emission position r s and measure the first sound pressure u r at n first sound pressure measurement positions r obs (t, r r , r s ) after the emission time t;

[0013] Step S11: Establish a scanning surface structure simulation model of the first container filled with inhomogeneous fluid, discretize the scanning surface structure simulation model, add the wave rules after discretization, and configure the sound speed of each discrete point to obtain a scanning surface sound speed field model

[0014] Step S12: Use the gradient optimization algorithm to iteratively correct the scanning surface sound speed field model to obtain a scanning surface sound speed field model that fits the sound speed field within the scanning surface in step S10

[0015] Step S13: Use the scanning surface sound speed field model to calculate the shortest propagation duration tof between any two points inside it. All the shortest propagation durations between two points in the scanning surface sound speed field model constitute the ultrasonic simulation model.

[0016] Preferably, let k ≥ 0. In step S12, the method of using the gradient optimization algorithm to iteratively correct the scanning surface sound speed field model includes the following steps:

[0017] Step S120. Before the (k + 1)-th iteration, simulate all the transmitting and receiving processes of Step S10 in the scanning plane sound speed field model . The scanning plane sound speed field model outputs the second sound pressure at the corresponding position at the measured post-transmission duration t

[0018] Let

[0019]

[0020] If , then terminate the iteration and obtain the scanning plane sound speed field model If , then perform the (k + 1)-th iteration; ξ is the maximum threshold of the second sound pressure and the first sound pressure

[0021] Step S121. At the (k + 1)-th iteration, the gradient operator

[0022] g k+1 = J T Δu k (2) where J represents the Jacobian matrix Δu k is the difference between the second sound pressure and the first sound pressure output by the scanning plane sound speed field model ;

[0023] Use the gradient optimization algorithm to find the search direction and the search step size, and update the scanning plane sound speed field model

[0024]

[0025] In the formula, α k+1 is the iteration step size, d k+1 is the search direction, d k+1 = f(g k+1 );

[0026] Let k = k + 1, and continue to execute Step S120

[0027] Further preferably, in Step S121, d k+1 = -g k+1 , and α k+1 is the iteration step size obtained by using the line search method

[0028] Further preferably, in Step S10, in each transmitting and receiving process, transmit the ultrasonic wave in the first shooting direction from the first transmitting position, and the first transmitting position, the first shooting direction, and n first sound pressure measurement positions are all set on the scanning plane; in Step S120, in the scanning plane sound speed field model ν k pThe method for simulating one transmission and reception process of step S10 is as follows: taking the relative positions of the first transmission position, the n first sound pressure measurement positions, and the container filled with inhomogeneous fluid in the one transmission and reception process of step S10 as the reference basis, determining the second transmission position and the n second sound pressure measurement positions respectively according to the position of the scanning plane structure simulation model, setting the parameters of the second ultrasonic wave simulatedly transmitted at the second transmission position, the second ultrasonic wave simulating the ultrasonic wave, and the scanning plane sound velocity field model Measure the second sound pressure at the second sound pressure measurement position at time t after transmitting the second ultrasonic wave.

[0029] Preferably, the container filled with inhomogeneous fluid is a transformer outgoing line device, and in step S20, n > 31.

[0030] Preferably, in step S11, set the maximum grid size Δs for discretization processing, and perform discretization processing on the scanning plane structure simulation model;

[0031] The wave rule after the discretization processing is

[0032]

[0033] where u(x, z, t) is the sound pressure field, (x, z) are respectively the abscissa and ordinate of the discrete points in the scanning plane, t is the time, v(x, z) is the sound velocity at the discrete point (x, z), Δt is the calculation time step, Δs is the maximum grid size, M represents 0.5 times the order of difference accuracy, and C is the difference coefficient.

[0034] Further preferably, in step S11, the maximum grid size Δs is 1 / 8 of the minimum wavelength of the ultrasonic wave; the calculation time step Δt satisfies v min is the minimum sound velocity of the medium in the detection plane.

[0035] Further preferably, in step S13, assume that the discrete point p is within the previous sound propagation depth, corresponding to the position (x p , z p ); the point q is within the subsequent sound propagation depth, corresponding to the position (x q , z q ), then the virtual propagation duration of the two discrete points p and q is

[0036]

[0037] In the formula, ν(x p , z p ) is the sound velocity of the discrete point (x p , z p );

[0038] If point p and point q are within the same sound propagation depth, or point p and point q are not within two adjacent sound propagation depths, or point p and point q are within two adjacent sound propagation depths, but point p is in the subsequent sound propagation depth and point q is in the previous sound propagation depth, then let tof pq = ∞, indicating that sound waves cannot propagate between p and q for the time being;

[0039] For point p and point q not within two adjacent sound propagation depths, and the sound propagation depth to which point p belongs is before the sound propagation depth to which point q belongs, the shortest path search algorithm is used to measure the shortest sound propagation duration between point p and point q, and tof is updated pq .

[0040] More preferably, in step S13, for point p and point q not within two adjacent sound propagation depths, and the sound propagation depth to which point p belongs is before the sound propagation depth to which point q belongs, the Viterbi shortest path search algorithm is used to measure the shortest sound propagation duration between point p and point q.

[0041] Preferably, in step S20, an ultrasonic transmitter is used to transmit ultrasonic waves. The ultrasonic transmitter is a linear array ultrasonic transmitter composed of m ultrasonic transmitting units. An ultrasonic receiver is used to measure the sound pressure at n third sound pressure measurement positions. The ultrasonic receiver is a linear array ultrasonic receiver composed of n ultrasonic receiving units, where n ≥ m > 2.

[0042] Preferably, in step S21, the full focus imaging algorithm is

[0043]

[0044] where I(x,z) is the pixel value at the discrete point (x,z), h( ) is the Hilbert transform, m is the number of third emission positions, n is the number of third sound pressure measurement positions, u ij is the third sound pressure when the ultrasonic wave emitted from the i-th third emission position propagates to the j-th third sound pressure measurement position, t i (x,z) represents the propagation time when the ultrasonic wave emitted from the i-th third emission position propagates to the position (x,z), and t j (x,z) represents the propagation time when the ultrasonic wave propagates from the discrete point (x,z) position to the j-th third sound pressure measurement position.

[0045] A computer-readable medium storing an ultrasonic imaging program, which, when executed by a processor, implements the ultrasonic imaging method for a container filled with an inhomogeneous fluid as described above.

[0046] An ultrasonic detector, comprising a fixing frame, m ultrasonic transmitting units, n ultrasonic receiving units, and a processor. It is characterized in that it further includes the aforementioned computer-readable medium storing an ultrasonic imaging program. The ultrasonic transmitting units are used to transmit a fourth ultrasonic wave identical to the ultrasonic wave. The m ultrasonic transmitting units, the n ultrasonic receiving units, and the fixing frame are fixedly connected, and the fourth ultrasonic waves of the m ultrasonic transmitting units are made to be shot towards and the sound pressure measurement positions of the n ultrasonic receiving units are arranged on the same plane.

[0047] The beneficial effects of the present invention are as follows:

[0048] 1. In the ultrasonic imaging method for a container filled with an uneven fluid of the present invention: Step S20 is to obtain the third emission position, the third shooting direction, the n third sound pressure measurement positions during m emission and reception processes carried out within the detection plane of the second container filled with an uneven fluid, and the third sound pressure at the third sound pressure measurement position at the time t after emission; in Step S21, using the ultrasonic simulation model corresponding to the detection plane of the second container filled with an uneven fluid and the full-focus imaging algorithm, and combining the data obtained in Step S21, the pixel values of all points within the ultrasonic simulation model are obtained; in Step S22, by outputting the pixel values of all points within the ultrasonic simulation model, ultrasonic imaging of the medium within the detection plane of the second container filled with an uneven fluid can be achieved. In the construction method of the ultrasonic simulation model,

[0049] Step S10 is to obtain the measured data of scanning the detection plane of the first container filled with an uneven fluid using a two-dimensional sound field arranged within the detection plane, so as to obtain, in Step S11 and Step S12, a scanned surface sound velocity field model having the same ultrasonic sound velocity characteristics as the medium within the detection plane of the first container filled with an uneven fluid Step S11 is to establish a scanned surface sound velocity field model of the first container filled with an uneven fluid Step S12 is to optimize the scanned surface sound velocity field model So that the obtained scanned surface sound velocity field model Fits the sound velocity field within the scanned surface of Step S10. Step S13 is to use the obtained scanned surface sound velocity field model To determine the shortest ultrasonic propagation duration between any two points in the scanned surface sound velocity field model, and all the shortest propagation durations between two points obtained constitute the ultrasonic simulation model corresponding to the detection plane, which can thus be applied to Step S21. The ultrasonic simulation model constructed using this ultrasonic simulation model construction method solves the problem that it is difficult to measure the sound velocity characteristics of an uneven fluid medium, and when applied to ultrasonic imaging of a container filled with an uneven fluid, it can improve the imaging accuracy.

[0050] 2. In the ultrasonic imaging method for a container filled with non-uniform fluid of the present invention: in the step S12, the use of a gradient optimization algorithm can improve the optimization efficiency of the sound velocity field model.

[0051] 3. In the ultrasonic imaging method for a container filled with non-uniform fluid of the present invention: in the step S121, d k+1 = -g k+1 , α k+1 is the iteration step size obtained by using the line search method. It corresponds to the steepest gradient descent optimization algorithm, with a fast convergence speed and high optimization efficiency.

[0052] 4. In the ultrasonic imaging method for a container filled with non-uniform fluid of the present invention: in the step S10, during each transmission and reception process, the ultrasonic wave is transmitted from 1 of the first transmission positions in the first transmission direction. Compared with the situation after transmitting ultrasonic waves from more than 2 first transmission positions, the sound pressure at the first sound pressure measurement position will not be superimposed, and the effect of each ultrasonic wave in the detection plane can be distinguished. Similarly, during each simulated transmission and reception process in the step S120, the ultrasonic wave is transmitted from 1 of the second transmission positions in the second transmission direction. The first transmission position, the first transmission direction, and n first sound pressure measurement positions are all set on the scanning plane, and a two-dimensional sound field can be formed in the detection plane.

[0053] 5. In the ultrasonic imaging method for a container filled with non-uniform fluid of the present invention: after the transformer is filled with insulating oil, it belongs to a container filled with non-uniform fluid. During actual use, the consistency of the insulating oil volume in the transformer is relatively good, so the reproducibility is high. Due to the large size of the transformer outlet device, n > 31, the ultrasonic wave reception range can be effectively increased, the calculation area can be increased, and the detection efficiency can be improved; at the same time, the data volume can be increased, and there can be more reference quantities during the inversion calculation of the sound velocity field model, improving the fitting degree of the sound velocity field model.

[0054] 6. In the ultrasonic imaging method for a container filled with non-uniform fluid of the present invention: the existing wave equation is applicable to continuous media, while the ultrasonic propagation characteristics of the media in a container filled with non-uniform fluid are inconsistent. Only after discretization processing can the sound velocity be configured for different media on this basis. Add the discretized wave equation to the scanning plane structure simulation model, configure the initial sound velocity of each discrete point in the scanning plane structure simulation model, and obtain the scanning plane sound velocity field model. In this way, an initial scanning plane sound velocity field model simulating the sound propagation characteristics is constructed to facilitate optimization fitting.

[0055] 7. In the ultrasonic imaging method of the container filled with non-uniform fluid according to the present invention: In the step S11, when discretizing the model space, take 1 / 8 of the minimum wavelength of the ultrasonic wave as the maximum grid size Δs, so as to finely distinguish the media on the ultrasonic wave propagation path. Calculate the time step Δt to satisfy v min is the minimum sound speed of the medium in the detection plane. When the obtained discrete points are used for finite difference forward simulation calculation, it can effectively weaken the physical dispersion phenomenon of the sound wave during propagation while ensuring the stability of the difference format.

[0056] 8. In the ultrasonic imaging method of the container filled with non-uniform fluid according to the present invention: In the step S13, since the grid size is smaller than the wavelength, the grids at the same depth position are approximately considered as isotropic media and the sound speed does not change. Therefore,

[0057] Let the discrete point p be within the previous sound propagation depth, corresponding to the position (x p , z p ); the point q is within the subsequent sound propagation depth, corresponding to the position (x q , z q ), then the virtual propagation time

[0058]

[0059] In the formula, v(x p , z p ) is the sound speed of the discrete point (x p , z p );

[0060] If the point p and the point q are within the same sound propagation depth, or the point p and the point q are not within two adjacent sound propagation depths, or the point p and the point q are within two adjacent sound propagation depths, but the point p is within the subsequent sound propagation depth and the point q is within the previous sound propagation depth, then let tof pq = ∞, indicating that the sound wave cannot propagate between p and q temporarily;

[0061] For the point p and the point q that are not within two adjacent sound propagation depths, and the sound propagation depth to which the point p belongs is before the sound propagation depth to which the point q belongs, then use the shortest path search algorithm to measure the shortest sound propagation time between the point p and the point q, and update tof pq .

[0062] 9. In the ultrasonic imaging method of a container filled with non-uniform fluid according to the present invention: Ultrasonic waves are emitted using an ultrasonic wave transmitter, which is a linear array ultrasonic wave transmitter composed of m ultrasonic wave transmitting units. The ultrasonic wave receiver measures the sound pressure at n third sound pressure measurement positions, and the ultrasonic wave receiver is a linear array ultrasonic wave receiver composed of n ultrasonic wave receiving units, where n ≥ m > 2. This can improve the data acquisition efficiency.

[0063] 10. In the computer-readable medium storing the ultrasonic imaging program according to the present invention, after the ultrasonic imaging program is executed by a processor, it implements the ultrasonic imaging method of the container filled with non-uniform fluid described above. Its reproduction method has good consistency, high calculation efficiency, and fast imaging speed.

[0064] 11. The ultrasonic detector according to the present invention includes a fixing frame, m ultrasonic wave transmitting units, n ultrasonic wave receiving units, a processor, and the computer-readable medium storing the ultrasonic imaging program described above. The m ultrasonic wave transmitting units, the n ultrasonic wave receiving units, and the fixing frame are fixedly connected, and the fourth ultrasonic wave emission direction of the m ultrasonic wave transmitting units and the sound pressure measurement positions of the n ultrasonic wave receiving units are arranged on the same plane. Using it, ultrasonic detection imaging can be performed on the second container filled with non-uniform fluid. Description of the Drawings

[0065] Figure 1 It is a flowchart of an ultrasonic imaging method for a first container filled with non-uniform fluid.

[0066] Figure 2 It is a search schematic diagram of the propagation path of the first ultrasonic wave in an ultrasonic imaging method for a first container filled with non-uniform fluid. Detailed Embodiments

[0067] The present invention will be described below in the form of embodiments in conjunction with the drawings to assist those skilled in the art in understanding and implementing the present invention. Unless otherwise specified, the following embodiments and the technical terms therein should not be understood without the technical knowledge background of the present technical field.

[0068] The ultrasonic detector includes an ultrasonic wave generator, an ultrasonic wave transmitter, and an ultrasonic wave receiver.

[0069] The ultrasonic wave generator converts the alternating current energy from the power supply into electrical energy suitable for powering the transducer at the ultrasonic frequency, and is used to send high-voltage electrical pulses to the ultrasonic transducer.

[0070] An ultrasonic transducer converts energy from electrical energy to mechanical energy, or vice versa. This conversion can be accomplished through piezoelectric ceramics or magnetostrictive materials. According to function, it can be divided into an ultrasonic transmitter, an ultrasonic receiver, and an ultrasonic transceiver. The ultrasonic transmitter converts a high-voltage electrical pulse into ultrasonic vibrations. The ultrasonic receiver converts ultrasonic vibrations into electrical signals. The ultrasonic transceiver can transmit and receive ultrasonic waves.

[0071] Embodiment 1: An ultrasonic imaging method for a container filled with non-uniform fluid, comprising the following steps:

[0072] Step S10: Use a two-dimensional sound field to scan a first container filled with non-uniform fluid. Let the plane where the two-dimensional sound field is located be the scanning plane; in each transmission and reception process, transmit a first ultrasonic wave in a first transmission direction from a first transmission position, measure the sound pressure at n first sound pressure measurement positions respectively, where n≥2, and obtain the relative positions of the first transmission position, the first transmission direction, the n first sound pressure measurement positions and the first container filled with non-uniform fluid corresponding to each transmission and reception process, as well as the first sound pressure corresponding to the first sound pressure measurement position.

[0073] In the field of ultrasonic testing, the two-dimensional sound field means that the first transmission position, the first transmission direction, and the n first sound pressure measurement positions are all within the scanning plane. When using a two-dimensional sound field to scan a first container filled with non-uniform fluid, generally within the same scanning plane, it is necessary to move the first transmission position so that the first transmission position surrounds the four sides of the scanning plane to complete the scanning of the scanning plane; after completing the scanning of one scanning plane, stepwise move the scanning plane along the normal direction of the scanning plane to complete the slice scanning of the first container filled with non-uniform fluid.

[0074] In this embodiment, it is used to obtain the ultrasonic simulation model of the transformer outlet device. Therefore, the first container filled with non-uniform fluid to be scanned in this step is a transformer outlet device that can be used as a measurement standard, that is, a normal transformer outlet device, different from a faulty transformer outlet device.

[0075] Specifically, in ultrasonic testing, the ultrasonic feedback signal reflected by the medium of the third ultrasonic wave received by the ultrasonic receiving unit. Therefore, the ultrasonic receiver and the ultrasonic transmitter are arranged on the same side of the transformer outlet device, and the ultrasonic transmitter, the ultrasonic receiver and the stepwise displacement mechanism are fixedly connected, so that the stepwise displacement mechanism drives the ultrasonic transmitter and the ultrasonic receiver to perform stepwise scanning on the shell of the transformer outlet device in a normal state.

[0076] The ultrasonic transmitter can be composed of 1 ultrasonic transmitting unit. However, in practice, to improve efficiency, the ultrasonic transmitter usually consists of a linear array of more than 2 ultrasonic transmitting units. In this embodiment, a two-dimensional sound field detection transformer outlet device is used. Therefore, a linear array ultrasonic transmitter is selected for the ultrasonic transmitter, which is composed of more than two ultrasonic transmitting units, and a linear array ultrasonic transmitter is selected for the ultrasonic receiver, which is composed of n ultrasonic receiving units. When in use, the ultrasonic transmitting unit, the ultrasonic wave emitted by the ultrasonic transmitting unit, and the n ultrasonic receiving units should all be arranged on the same plane.

[0077] In the ultrasonic detectors on the market, generally, the functions of the ultrasonic transmitting unit and the ultrasonic receiving unit are realized through an ultrasonic transceiver unit, which is smaller in volume and easier to control for scanning operations. At this time, the number of the ultrasonic transmitting unit and the ultrasonic receiving unit is the same. When scanning the housing of the transformer outlet device, it is best to make the ultrasonic transmitter closely adhere to the surface of the housing of the transformer outlet device.

[0078] Taking the use of an ultrasonic transceiver with a linear array arrangement of 16 ultrasonic transceiver units as an example, when in use, at each scanning position, one of the ultrasonic transceiver units is used to emit a first ultrasonic wave (the first ultrasonic wave is generally a pulsed ultrasonic wave). This first ultrasonic wave is reflected in the transformer outlet device, and the ultrasonic feedback signal returns to the ultrasonic receiving unit. All the ultrasonic receiving units wait to receive the ultrasonic feedback signal. After all the ultrasonic receiving units receive the ultrasonic feedback signal, a single transmission and reception process is completed. Then, the ultrasonic transceiver unit is replaced to emit another first ultrasonic wave and the next transmission and reception process is executed until all 16 ultrasonic transceiver units have completed a single transmission and reception process, at which point the scanning of this position is completed; the ultrasonic transceiver is stepped to the next scanning position to scan the new scanning position until the scanning of the entire scanning surface is completed around the perimeter of the scanning surface; after the scanning of one scanning surface is completed, the scanning surface is stepped along the normal direction of the scanning surface to complete the sliced scanning of the first container filled with non-uniform fluid.

[0079] During each transmission and reception process, the relative positions of the ultrasonic transmitter, the ultrasonic receiver, and the transformer outgoing line device are determined. Among them, the ultrasonic excitation point of the ultrasonic transmitting unit corresponds to the first transmission position, the ultrasonic wave emitted by the ultrasonic transmitting unit is directed towards the first direction, and the position of the ultrasonic receiving unit corresponds to the first sound pressure measurement position. During each transmission and reception process, it is necessary to record the relative positions of the first transmission position, the first direction, the first sound pressure measurement position, and the first container filled with non-uniform fluid, as well as the transmitted sound pressure of the first ultrasonic wave and the first sound pressure received at each first sound pressure measurement position. Recording the relative positions of the first transmission position, the first direction, the first sound pressure measurement position, and the first container filled with non-uniform fluid, as well as the transmitted sound pressure of the first ultrasonic wave and the first sound pressure received at each first sound pressure measurement position during each transmission and reception process, is to reproduce this transmission and reception process in subsequent simulation.

[0080] When establishing a simulation model of the transformer outgoing line device in the model space of the sound propagation simulation software and performing sound propagation simulation, when the relative positions of the second transmission position, the second direction, n second sound pressure measurement positions, and the simulation model of the transformer outgoing line device are the same as those of the first transmission position, the first direction, n first sound pressure measurement positions, and the transformer outgoing line device, the second ultrasonic wave is simulated to be emitted along the second direction at the second transmission position. The second ultrasonic wave should simulate the first ultrasonic wave, that is, the physical quantities such as the transmitted sound pressure and waveform of the second ultrasonic wave should be the same as those of the first ultrasonic wave, and the second sound pressure at the n second sound pressure measurement positions that match is calculated through forward simulation. Generally, in the sound propagation simulation software, only by inputting relevant parameters can the second sound pressure at the second sound pressure measurement position be obtained.

[0081] When simulating the emission of the second ultrasonic wave, if the second ultrasonic wave propagates spherically, that is, it spreads diffusely in all directions within the scanning plane, it is not necessary to make the second direction the same as the first direction.

[0082] In the simulation, if the difference between the second sound pressure at the second sound pressure measurement position and the first sound pressure at the equivalent first sound pressure measurement position is within the sound pressure attenuation equivalent threshold, it is considered that the simulated propagation path of the second ultrasonic wave emitted from the second emission position in the model space to the second sound pressure measurement position is the same as the measured propagation path of the first ultrasonic wave emitted from the first emission position to the corresponding first sound pressure measurement position in this step, reproducing the emission and reception process from the first emission position to the first sound pressure measurement position. The ultrasonic propagation characteristics of the medium on this simulated propagation path are the same as those on this measured propagation path. In the simulation, if the emission and reception processes from the first emission position to the first sound pressure measurement position are reproduced for all sound pressure detection points corresponding to the same second ultrasonic wave, that is, the simulated propagation path of the second ultrasonic wave measured in the model space is the same as the actual propagation path of the first ultrasonic wave in this step. In other words, the ultrasonic sound velocity characteristics of the medium on the n simulated propagation paths of the second ultrasonic wave of the transformer outlet device simulation model in the model space are the same as those of the medium on the n measured propagation paths of the first ultrasonic wave of the actual transformer outlet device. By traversing and simulating the scanning of the transformer outlet device simulation model, when the emission and reception processes of the first emission position in this step are reproduced at all second emission positions, it is considered that the ultrasonic sound velocity characteristics of the transformer outlet device simulation model in the model space are the same as those of the measured transformer outlet device. In this way, a sound velocity field model of the transformer outlet device is obtained.

[0083] During actual operation, generally, the relative positions of the first emission position, the first emission direction, the first sound pressure measurement position, and the transformer outlet device are kept in a pattern that is easy to reproduce, so as to reduce the complexity of the simulation.

[0084] After obtaining the sound velocity field model of the first container filled with non-uniform fluid, according to Fermat's principle, the shortest propagation time of flight tof between any two points p and q in the model space is determined using the shortest path search algorithm. pq , this shortest propagation time of flight tof pq corresponds to the propagation path of the second ultrasonic wave. The shortest propagation times of flight between all pairs of points can be used to construct the ultrasonic simulation model of the first container filled with non-uniform fluid.

[0085] During use, perform m emission and reception processes on the outgoing line device of the transformer under test. During each emission and reception process, emit the third ultrasonic wave in the third emission direction at the third emission position, and measure the third sound pressure at n third sound pressure measurement positions respectively. The emission sound pressure, waveform, etc. of the third ultrasonic wave are the same as those of the first ultrasonic wave. When the housing of the outgoing line device of the transformer under test coincides with the housing of the outgoing line device in this step, the detection plane determined by the third emission direction and the n third sound pressure measurement positions is parallel to the scanning plane in this step. Input the relative positions of the third emission position, the third emission direction, the n third sound pressure measurement positions, and the first container filled with non-uniform fluid obtained from the m emission and reception processes into the ultrasonic simulation model, as well as the third sound pressure corresponding to the third sound pressure measurement position. The ultrasonic simulation model outputs the propagation time t of the third ultrasonic wave emitted in the third emission direction at the i-th third emission position to the position of the discrete point (x, z). i (x, z) and the propagation time t of the third ultrasonic wave from the position of the discrete point (x, z) to the j-th third sound pressure measurement position j (x, z), where m ≥ 1; i takes 1, 2, 3,..., m respectively; j takes 1, 2, 3,..., n respectively; use the imaging algorithm to calculate the pixel values of the discrete points in the detection plane, and obtain the ultrasonic image in the detection plane of the second container filled with non-uniform fluid.

[0086] When emitting the third ultrasonic wave, if the third ultrasonic wave propagates in a hemispherical shape, that is, it spreads diffusely in all directions in the scanning plane, it is not necessary to make the third emission direction the same as the first emission direction.

[0087] In the ultrasonic detectors on the market, the width of the ultrasonic transceiver unit is about 2 cm. Therefore, the distance between two adjacent ultrasonic transceiver units is about 5 cm. Combining the principle of the present invention, it can be seen that:

[0088] First, theoretically, the closer the distance between adjacent first emission positions and the closer the distance between adjacent first sound pressure measurement positions, the higher the inversion accuracy of the ultrasonic sound velocity characteristics of the sound velocity field model of the outgoing line device of the transformer and the ultrasonic sound velocity characteristics of the actual outgoing line device of the transformer. However, the coverage range of a single emission and reception process is small and the scanning efficiency is low. Therefore, it is necessary to balance the scanning efficiency and the fitting accuracy, that is, it is necessary to appropriately control the distance between two adjacent first emission positions and the distance between two adjacent first sound pressure measurement positions.

[0089] Second, theoretically, during a single emission and reception process, the larger n is, the more reference points for inversion calculation are obtained, thereby improving the similarity between the ultrasonic sound velocity characteristics of the simulation model of the outgoing line device of the transformer in the model space and the ultrasonic sound velocity characteristics of the actual outgoing line device of the transformer.

[0090] Thirdly, the n first sound pressure measurement positions do not need to be regularly arranged either. However, the array setting of the n first sound pressure measurement positions in cooperation with the step-by-step scanning is more in line with the grid measures of the simulation, and can also avoid missing scanning positions during scanning. Moreover, this can reduce the model complexity and operation difficulty.

[0091] Since the size of the transformer is much larger than the sizes of the ultrasonic transmitting unit and the ultrasonic receiving unit, theoretically, the more the number of the ultrasonic transmitting unit and the ultrasonic receiving unit, the better. The more the number of the ultrasonic receiving unit, the wider the coverage range of the first ultrasonic wave. The more the number of the ultrasonic transmitting unit, the fewer the number of times of step-by-step movement required, which can improve the transformer detection efficiency. In this embodiment, n is preferably ≥ 32.

[0092] Currently, the COMSOL software can use the finite element method to realize the computer simulation of the acoustic wave propagation process. In this embodiment, the finite difference method is used for self-programming to realize the forward simulation process. When simulating the acoustic wave propagation process, it is necessary to establish the initial sound velocity field model of the transformer outgoing line device, and then continuously correct the sound velocity field model of the transformer outgoing line device according to the inversion process, so that the second sound pressure received in the forward simulation process approaches the first sound pressure actually collected.

[0093] Step S11: Establish a simulation model of the first container filled with inhomogeneous fluid in the model space of the acoustic propagation simulation software, discretize the model space, use the parallel plane of the scanning plane to cut the simulation model layer by layer, add the wave rules after discretization processing, and configure the sound velocity for each discrete point on the obtained space slice to obtain the initial sound velocity field model of the model space;

[0094] When discretizing the model space of the acoustic propagation software, it is necessary to reasonably select the grid size and the calculation time step. The method of selecting the grid size and the calculation time step is as follows:

[0095] Take 1 / 8 of the minimum wavelength as the maximum grid size of the model space, and the calculation time step satisfies the Courant-Friedrichs-Lewy (CFL) condition, that is:

[0096]

[0097] where v min is the minimum sound velocity value in the model space, Δt is the calculation time step, and Δs is the maximum grid size of the model space;

[0098] According to the minimum sound velocity value v min in the model space, calculate the minimum wavelength, and then determine the maximum grid size in the model space. Then, combined with Equation (6), the range of the calculation time step can be determined.

[0099] By simulating the acoustic wave propagation process by computer, the received sound pressure signal, the wave source, and the propagation medium satisfy the wave equation:

[0100]

[0101] where u(x,z,t) is the sound pressure field, (x,z) are the horizontal and vertical coordinates of the medium in the two-dimensional acoustic field respectively, t is the time, v(x,z) is the sound speed in the medium at the point (x,z), is the Laplace operator, and s(x,z,t) is the source term.

[0102] According to the finite difference method, the discrete form converted from this wave equation is:

[0103]

[0104] where u(x,z,t) is the sound pressure field, (x,z) are the horizontal and vertical coordinates of the medium in the spatial slice respectively, t is the time, v(x,z) is the sound speed in the medium at the discrete point (x,z), Δt is the calculation time step, Δs is the maximum grid size of the model space, M represents 0.5 times the order of the difference accuracy, and C is the difference coefficient. For example, M = 1 represents 2nd-order difference accuracy; M = 2 represents 4th-order difference accuracy; M = 3 represents 6th-order difference accuracy; and so on...

[0105] Use parallel planes of the scanning plane to slice the simulation model layer by layer, that is: when the simulation model of the first container filled with non-uniform fluid coincides with the first container filled with non-uniform fluid in step S10, the cutting plane in this step is parallel to the scanning plane in step S10. The purpose is to make the setting method of the scanning plane of the first container filled with non-uniform fluid in step S10 the same as the setting method of the cutting plane of the simulation model.

[0106] Add the discretized wave equation in the model space, input the initial sound speed of the medium at each discrete point in the model space, and obtain the initial sound speed field model of the model space. For the initial sound speed of the medium in the simulation model of the transformer outlet device, the initial sound speed of its shell is the theoretical sound speed at normal temperature and pressure, the initial sound speed of its internal insulating oil is the theoretical sound speed at normal temperature and pressure without disturbance, and the sound speed of the medium at discrete points outside the transformer outlet device in the model space is the theoretical sound speed of air at normal temperature and pressure.

[0107] Step S12: Forward simulate the transmitting and receiving process of step S10 at the matching position of the simulation model. Targeting that the second sound pressure measured at the matching position of the simulation model approaches the first sound pressure obtained in step S10, use the gradient optimization algorithm to iteratively correct the initial sound velocity field model, so that the ultrasonic sound velocity properties in the model space fit the ultrasonic sound velocity properties of the first container filled with non-uniform fluid in step S10, and obtain the final sound velocity field model of the model space.

[0108] Use acoustic propagation simulation software to simulate the emission of the second ultrasonic wave along the second radiation direction at the matching second emission position, and forward simulate and calculate the second sound pressure at the matching n second sound pressure measurement positions. Among them, when the relative positions of the second emission position, the second radiation direction, the n second sound pressure measurement positions and the transformer outgoing line device simulation model are the same as those of the first emission position, the first radiation direction, the n first sound pressure measurement positions and the transformer outgoing line device in step S10, the second emission position matches the first emission position, the second radiation direction matches the first radiation direction, and the second sound pressure measurement position matches the first sound pressure measurement position. The acoustic characteristics such as the emission sound pressure and waveform of the second ultrasonic wave are the same as those of the first ultrasonic wave in step S10.

[0109] Calculate the data difference between the ultrasonic feedback signal obtained from the forward numerical simulation process and the actually collected ultrasonic feedback signal, which is expressed as the objective function in the sense of least squares:

[0110]

[0111] where u cal is the second sound pressure calculated by forward simulation in the sound velocity field v, and u obs is the first sound pressure.

[0112] The discretized objective function is:

[0113]

[0114] where u cal (t, r r , r s |ν) is the second sound pressure calculated by forward simulation in the sound velocity field v, u obs (t, r r , r s ) is the first sound pressure, r s represents the second emission position, and r t represents the second sound pressure measurement position.

[0115] The definition of the gradient in the model is the partial derivative of the objective function with respect to the sound velocity field in the model, and the gradient operator is:

[0116]

[0117] Among them, J represents the Jacobian matrix, Δu k is the sound speed field model the difference between the second sound pressure and the first sound pressure output.

[0118] Combined with the parameter information in the model space, such as the sound speed, the second sound pressure, and the first sound pressure at each discrete point, calculate the gradient of the objective function with respect to the sound speed field at all positions at any time in the model space.

[0119] Based on the gradient optimization algorithm, minimize the objective function and iteratively update the sound speed field model. The flow chart is as Figure 1 shown. The specific process is as follows:

[0120] Starting from the given initial sound speed field model v0 for calculation, since the relationship between the data and the model parameters is non-linear, multiple iterations are required to converge to the local minimum of the objective function near the sound speed field model.

[0121] In each iteration process, find the search direction and the search step size through the gradient optimization algorithm, and update the sound speed field model. Taking the (k + 1)-th iteration as an example:

[0122] v k+1 = v k + α k+1 d k+1 (12)

[0123] Among them, ν k+1 is the sound speed field model at the (k + 1)-th iteration, α k+1 is the iteration step size of the (k + 1)-th iteration, d k+1 is the search direction of the (k + 1)-th iteration, and d k+1 = f(g k+1 ).

[0124] Since different gradient descent methods are used, the f(g) function is also different. In this embodiment, the steepest gradient descent optimization algorithm is used. Therefore, d k+1 = -g k+1 . In other embodiments, the conjugate gradient method, the spectral conjugate gradient method, the hybrid conjugate gradient method, etc. can also be used.

[0125] When E(v k+1 ) < E(v k ) it means that the objective function value decreases until E(v k+1 ) ≤ ξ. At this time, the sound speed field model v k+1 is the final sound speed field model; where ξ is the maximum threshold of the difference between the second sound pressure obtained by the forward numerical simulation and the first sound pressure actually collected.

[0126] Step S13: Combine the discrete points and the final sound velocity field model to determine the shortest propagation time tof between any two points p and q on each spatial slice in the model space. pq All the shortest propagation times constitute an ultrasonic simulation model.

[0127] Based on Fermat's principle, that is, "ultrasound always propagates between two fixed points in space along the path with the shortest time", the propagation path of sound rays in the model space is approximately the shortest path problem in computer graph theory. According to the Viterbi shortest path search algorithm, calculate the sound propagation path and sound propagation time between any two fixed points in the model space, and establish a sound time matrix for the model space.

[0128] Taking a 4-layer space model as an example, its sound ray search is as Figure 2 shown. Among them, the x-axis direction is horizontal to the right, and the z-axis direction is the depth direction.

[0129]

[0130] Let the number of grid points at the same depth position after discretization be A, and the number of all discrete points be D. Then Figure 1 in it, A = 7.

[0131] After abstraction, there is the following discrete point matrix:

[0132]

[0133] Among them, D = (B + 1)·A.

[0134] Assume that the second emission position is S, the second sound pressure measurement position is set at point e, and the second ultrasonic wave is emitted from point S and reaches point e.

[0135] To construct the Viterbi shortest path search algorithm, it is necessary to establish a sound time matrix to represent the sound propagation time between any two fixed points. Since the grid size is smaller than the wavelength, the grids at the same depth position are approximately considered to be isotropic media and the sound velocity does not change. Therefore, connect the paths between all pairs of grid points, which are all the search paths from point S to point e. Use tof pq to represent the propagation time between any two points p and q. Only when p and q are in adjacent grids and not at the same depth position can sound waves propagate between p and q. Therefore, establish a sound time matrix for the model space:

[0136]

[0137] Among them, A represents the number of grid points at the same depth position after discretization, and D represents the number of all grid points.

[0138] When the discrete point p is within the propagation depth of the previous sound wave, corresponding to the position (x p , z p ) and the point q is within the propagation depth of the subsequent sound wave, corresponding to the position (x q , z q ), then the virtual propagation time of the two discrete points p and q

[0139]

[0140] where v(x p , z p ) is the sound speed of the discrete point (x p , z p ).

[0141] If the points p and q are within the same sound wave propagation depth, or the points p and q are not within two adjacent sound wave propagation depths, or the points p and q are within two adjacent sound wave propagation depths, but point p is within the subsequent sound wave propagation depth and point q is within the previous sound wave propagation depth, then let tof pq = ∞, indicating that the sound wave cannot propagate between p and q temporarily.

[0142] For the points p and q that are not within two adjacent sound wave propagation depths, and the sound wave propagation depth to which point p belongs is before the sound wave propagation depth to which point q belongs, the shortest path search algorithm is used to measure the shortest sound wave propagation time between points p and q, and tof pq is updated.

[0143] According to Fermat's principle, the sound wave propagation path is the path with the shortest sound wave propagation time among all possible paths. Therefore, the Viterbi shortest path search algorithm is used to calculate the sound wave propagation path. First, starting from the sound source point S, for the A discrete points in the first layer, obviously, the time between S and each point in the first layer is the shortest time. The sound travel time matrix is searched to obtain the time tof = tof 1a , where a = 1, 2, 3,..., A; then the propagation time of the A discrete points in the second layer is calculated. For a specific point b among them, its sound wave propagation time is tof 1b = min(tof 1a + tof ab ); through iterative calculation in this way, the shortest propagation time from the sound source point S to all grid points can be obtained, and finally the sound wave propagation path corresponding to the shortest time can be calculated according to the backtracking principle.

[0144] Step S14: Input the relative positions of the third emission positions, third emission directions, n third sound pressure measurement positions, and the second container filled with non-uniform fluid obtained from m emission and reception processes into the ultrasonic simulation model, as well as the third sound pressure corresponding to the third sound pressure measurement positions. The ultrasonic simulation model outputs the propagation time t of the third ultrasonic wave emitted in the i-th third emission position in the third emission direction to the position of the discrete point (x, z). i (x, z) and the propagation time t of the third ultrasonic wave from the position of the discrete point (x, z) to the j-th third sound pressure measurement position j (x, z), where m ≥ 1; i takes 1, 2, 3, …, m respectively; j takes 1, 2, 3, …, n respectively. When it is assumed that the second container filled with non-uniform fluid coincides with the first container filled with non-uniform fluid, the detection plane determined by the third emission direction and n third sound pressure measurement positions is parallel to the scanning plane in Step S10, and the third ultrasonic wave is the same as the first ultrasonic wave in Step S10. Use the imaging algorithm to calculate the pixel values of the discrete points in the detection plane, and obtain the ultrasonic image in the detection plane of the second container filled with non-uniform fluid.

[0145] In this step, the emission and reception process can be actually carried out. That is, in each emission and reception process, the third ultrasonic wave is emitted in the third emission direction at the third emission position. The emission sound pressure, waveform, etc. of the third ultrasonic wave should be the same as those of the first ultrasonic wave, and the third sound pressure is measured at n third sound pressure measurement positions respectively. The emission and reception process can also be simulated. It can also be the relative positions of the third emission positions, third emission directions, n third sound pressure measurement positions, and the second container filled with non-uniform fluid involved in this emission and reception process, as well as the third sound pressure at the third sound pressure measurement positions, which are directly obtained without emission.

[0146] In this embodiment, the propagation time t of the third ultrasonic wave emitted in the i-th third emission position in the third emission direction to the position of the discrete point (x, z) is obtained i (x, z) and the propagation time t of the third ultrasonic wave from the position of the discrete point (x, z) to the j-th third sound pressure measurement position j (x, z), where i takes 1, 2, 3, …, m respectively; j takes 1, 2, 3, …, n respectively. Calculate the focusing law of the first ultrasonic wave focused on any position in the detection plane in the model space through post-processing, and combine the full focusing imaging algorithm to calculate the coherent summation of all the third ultrasonic waves at m third emission positions and n third sound pressure measurement positions at the discrete points in the detection plane of the simulation model.

[0147] Therefore, the imaging algorithm is used as

[0148]

[0149] Wherein, I(x, z) is the pixel value at the discrete point (x, z), h() is the Hilbert transform, m is the number of the third emission positions, n is the number of the third sound pressure measurement positions, u ij is the third sound pressure that the third ultrasonic wave emitted from the i-th third emission position propagates to the j-th third sound pressure measurement position, t i (x, z) represents the propagation time that the third ultrasonic wave emitted from the i-th third emission position propagates to the discrete point (x, z), t j (x, z) represents the propagation time that the third ultrasonic wave propagates from the discrete point (x, z) to the j-th third sound pressure measurement position.

[0150] In Equation (14), if the ultrasonic transmitter has only one ultrasonic emission unit, then m = 1; if the ultrasonic transmitter has m ultrasonic emission units, during use, at one scanning position, the ultrasonic transmitter may not be moved, but the m emission-reception processes may be performed by changing the ultrasonic emission units.

[0151] Coherent superposition does not change the actual sound pressure at the actual target point position, but is only a virtual focusing processing method, which can increase the amplitude when imaging the target point and improve the resolution.

[0152] Therefore, the array ultrasonic full focusing imaging of the indirect outlet device structure of the UHV transformer (converter transformer) reactor can be realized. Through the full focusing imaging of the entire transformer interior (including insulating oil and outlet device structure), it is judged whether the position state of the outlet device is normal.

[0153] Based on the method for constructing the ultrasonic detection model of the container filled with non-uniform fluid in Embodiment 1, the ultrasonic imaging method of the container filled with non-uniform fluid of the present invention can be obtained, including:

[0154] Step S20: Obtain the third emission positions, n third sound pressure measurement positions during the m emission-reception processes performed in the detection plane of the second container filled with non-uniform fluid, and the third sound pressure at the third sound pressure measurement positions at the time t after emission, where m≥1 and n≥2; the third emission positions, the third emission directions and the n third sound pressure measurement positions here are all relative positions with respect to the second container filled with non-uniform fluid. The time t after emission starts from the moment when the ultrasonic wave is emitted from the third emission position during each emission process, and after the time t has elapsed, the third sound pressure at the third sound pressure measurement positions is respectively obtained.

[0155] Step S21, using the ultrasonic simulation model of the second container containing the non-uniform fluid corresponding to the detection plane and the data obtained in step S21 to calculate the propagation time of the ultrasonic wave emitted from the i-th third transmitting position to point P (point P is a point in the ultrasonic simulation model) during the m-times of transmission and reception and the propagation time of the ultrasonic wave from point P to the j-th third sound pressure measurement position, and using the full-focus imaging algorithm to calculate the pixel value of point P; traversing the points in the ultrasonic simulation model as point P, and obtaining the pixel values ​​of all points in the ultrasonic simulation model, wherein i=1,2,…,m; j=1,2,…,n;

[0156] Step S22, outputting the pixel values ​​of all points in the ultrasound simulation model;

[0157] The method for constructing the ultrasonic simulation model comprises the following steps:

[0158] Step S10: scanning the first container containing the non-uniform fluid in the detection plane, and in each transmitting and receiving process of scanning the first container containing the non-uniform fluid in the detection plane, at the first transmitting position r s The ultrasonic wave is emitted, and n first sound pressure measurement positions r are measured respectively at a time t after the emission. r The first sound pressure u obs (t,r r ,r s ); The ultrasonic wave emitted in this step should have the same acoustic properties as the ultrasonic wave emitted in step S20. The direction of the ultrasonic wave emitted in this step is the first direction, the first emission position, the first emission direction and the n first sound pressure measurement positions are all set in the detection plane, and the time after emission t is calculated from the moment when the ultrasonic wave is emitted from the first emission position during each emission process, and after the time t, the first sound pressure at the first sound pressure measurement position is obtained respectively.

[0159] Step S11, establishing a scanning surface structure simulation model of the first container containing the non-uniform fluid, discretizing the scanning surface structure simulation model, adding the wave rule after the discrete processing, configuring the sound velocity of each discrete point, and obtaining the scanning surface sound velocity field model

[0160] Step S12: using a gradient optimization algorithm to iteratively correct the scanning surface sound velocity field model Obtain a scanning surface sound velocity field model that fits the sound velocity field in the scanning surface of step S10

[0161] Step S13: Using the scanning surface sound velocity field model Calculate the shortest propagation time tof between any two points in it, the scanning surface sound velocity field model The shortest propagation durations between all pairs of points therein constitute the ultrasonic simulation model;

[0162] Among them, the first container filled with non-uniform fluid can be the second container filled with non-uniform fluid, or a container equivalent to the second container filled with non-uniform fluid.

[0163] In the ultrasonic imaging method of the container filled with non-uniform fluid of the present invention, since the detection plane is determined, therefore, in the construction method of the ultrasonic simulation model, only steps S10 to S13 are required to construct the ultrasonic simulation model corresponding to the detection plane of the first container filled with non-uniform fluid, and the ultrasonic imaging method of the container filled with non-uniform fluid of the present invention can be realized. However, in practice, usually step S14 is added after step S13 to construct the ultrasonic simulation model of the first container filled with non-uniform fluid. In this way, when performing step S21, only the scanning plane ultrasonic simulation model of the first container filled with non-uniform fluid corresponding to the detection plane of step S20 needs to be selected, and the ultrasonic imaging method of the container filled with non-uniform fluid of the present invention can be realized.

[0164] Step 14 is: stepwise move the scanning plane in the normal direction of the detection plane, repeat steps S10 to S13, obtain the scanning plane ultrasonic simulation models on each scanning plane of the first container filled with non-uniform fluid, and align all the scanning plane ultrasonic simulation models in sequence to obtain the ultrasonic simulation model of the first container filled with non-uniform fluid.

[0165] Preferably, let k≥0. In step S12, use the gradient optimization algorithm to iteratively correct the scanning plane sound velocity field model The method includes the following steps:

[0166] Step S120: Before the (k + 1)-th iteration, simulate all the emission and reception processes of step S10 in the scanning plane sound velocity field model The scanning plane sound velocity field model Outputs the second sound pressure at the corresponding position when the measured emission duration t is reached

[0167] Let

[0168]

[0169] If Then terminate the iteration to obtain the scanning plane sound velocity field model If Then perform the (k + 1)-th iteration; ξ is the maximum threshold of the second sound pressure and the first sound pressure;

[0170] Step S121: At the (k + 1)-th iteration, the gradient operator

[0171] g k+1 = J T Δu k (16) where J represents the Jacobian matrix, Δu k is the difference between the second sound pressure and the first sound pressure output by the sound velocity field model of the scanning surface; Output the difference between the second sound pressure and the first sound pressure;

[0172] Use the gradient optimization algorithm to find the search direction and search step size, and update the sound velocity field model of the scanning surface

[0173]

[0174] In the formula, α k+1 is the iteration step size, d k+1 is the search direction, d k+1 = f(g k+1 );

[0175] Let k = k + 1, and continue to execute step S120.

[0176] Further preferably, in the step S121, d k+1 = -g k+1 , α k+1 is the iteration step size obtained by using the line search method.

[0177] Further preferably, in the step S10, during each transmission and reception process, the ultrasonic wave is transmitted from the first transmission position in the first radiation direction, and the first transmission position, the first radiation direction, and n first sound pressure measurement positions are all set on the scanning surface; in the step S120, the method of simulating one transmission and reception process of the step S10 in the sound velocity field model of the scanning surface is: based on the relative positions of the first transmission position, n first sound pressure measurement positions and the first container filled with non-uniform fluid in one transmission and reception process in the step S10 as a reference, according to the position of the scanning surface structure simulation model, determine the second transmission position and n second sound pressure measurement positions respectively, set the parameters of the second ultrasonic wave simulated to be transmitted at the second transmission position, the second ultrasonic wave simulates the ultrasonic wave, and the sound velocity field model of the scanning surface measures the second sound pressure at the second sound pressure measurement position at time t after transmitting the second ultrasonic wave. Among them, the acoustic properties of the ultrasonic wave transmitted in the step S20, the ultrasonic wave transmitted in the step S10, and the ultrasonic wave transmitted in the step S120 are the same. The second transmission position and n second sound pressure measurement positions are all relative positions to the position of the scanning surface structure simulation model.

[0178] Preferably, the container containing the non-uniform fluid is a transformer outgoing line device, and in the step S20, n > 31.

[0179] Preferably, in the step S11, a maximum grid size Δs for discretization processing is set, and the scanning surface structure simulation model is discretized.

[0180] The fluctuation rule after the discretization process is

[0181]

[0182] where u(x, z, t) is the sound pressure field, (x, z) are respectively the abscissa and ordinate of the discrete points in the scanning surface, t is the time, v(x, z) is the sound speed at the discrete point (x, z), Δt is the calculation time step, Δs is the maximum grid size, M represents 0.5 times the order of the difference accuracy, and C is the difference coefficient. Here, the time t is the same as the t in step S20, step S10, and step S120.

[0183] More preferably, in the step S11, the maximum grid size Δs is 1 / 8 of the minimum wavelength of the ultrasonic wave; the calculation time step Δt satisfies v min is the minimum sound speed of the medium in the detection plane.

[0184] More preferably, in the step S13, it is assumed that the discrete point p is within the previous sound propagation depth and corresponds to the position (x p , z p ); the point q is within the subsequent sound propagation depth and corresponds to the position (x q , z q ), then the virtual propagation duration of the two discrete points p and q is

[0185]

[0186] In the formula, v(x p , z p ) is the sound speed of the discrete point (x p , z p );

[0187] If the points p and q are within the same sound propagation depth, or, the points p and q are not within two adjacent sound propagation depths, or, the points p and q are within two adjacent sound propagation depths, but the point p is within the subsequent sound propagation depth and the point q is within the previous sound propagation depth, then let tof pq = ∞, indicating that the sound wave cannot propagate between p and q temporarily;

[0188] For points p and q not within two adjacent sound propagation depths, and the sound propagation depth to which point p belongs is before the sound propagation depth to which point q belongs, the shortest path search algorithm is used to measure the shortest sound propagation duration between points p and q, and tof is updated. pq 。

[0189] More preferably, in step S13, for points p and q not within two adjacent sound propagation depths, and the sound propagation depth to which point p belongs is before the sound propagation depth to which point q belongs, the Viterbi shortest path search algorithm is used to measure the shortest sound propagation duration between points p and q.

[0190] Preferably, in step S20, an ultrasonic transmitter is used to emit ultrasonic waves. The ultrasonic transmitter is a linear array ultrasonic transmitter composed of m ultrasonic transmitting units. An ultrasonic receiver is used to measure the sound pressure at n third sound pressure measurement positions. The ultrasonic receiver is a linear array ultrasonic receiver composed of n ultrasonic receiving units, where n ≥ m > 2.

[0191] Preferably, in step S21, the full focus imaging algorithm is

[0192]

[0193] where I(x,z) is the pixel value at the discrete point (x,z), h() is the Hilbert transform, m is the number of third emission positions, n is the number of third sound pressure measurement positions, u ij is the third sound pressure when the ultrasonic wave emitted from the i-th third emission position propagates to the j-th third sound pressure measurement position, and t i (x,z) represents the propagation time when the ultrasonic wave emitted from the i-th third emission position propagates to the (x,z) position, and t j (x,z) represents the propagation time when the ultrasonic wave propagates from the discrete point (x,z) position to the j-th third sound pressure measurement position.

[0194] A computer-readable medium storing an ultrasonic imaging program, which, when executed by a processor, implements the ultrasonic imaging method for a container filled with an inhomogeneous fluid described above.

[0195] An ultrasonic detector, including a fixing frame, m ultrasonic transmitting units, n ultrasonic receiving units, and a processor. It is characterized in that it further includes the computer-readable medium storing the ultrasonic imaging program described above. The ultrasonic transmitting units are used to emit fourth ultrasonic waves identical to the ultrasonic waves. The m ultrasonic transmitting units, the n ultrasonic receiving units, and the fixing frame are fixedly connected, and the fourth ultrasonic waves of the m ultrasonic transmitting units are directed towards the sound pressure measurement positions of the n ultrasonic receiving units and are arranged on the same plane.

[0196] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code. The solutions in the embodiments of the present application can be implemented using various computer languages. For example, object-oriented programming languages such as Java and interpreted scripting languages such as JavaScript, etc.

[0197] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in Figure 1 one or more of the flows Figure 1 or a plurality of flows and / or blocks

[0198] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device implements the functions specified in Figure 1 one or more of the flows Figure 1 or a plurality of flows and / or blocks

[0199] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one or more of the flows Figure 1 or a plurality of flows and / or blocks

[0200] Although the preferred embodiments of the present application have been described, those skilled in the art can make additional changes and modifications to these embodiments once they learn the basic creative concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications falling within the scope of the present application.

[0201] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application is also intended to include these modifications and variations.

[0202] The present invention has been described in detail above in conjunction with the drawings and embodiments. It should be understood that in practice, it is impossible to exhaustively describe all possible implementation manners, and here the inventive concept of the present invention is elaborated as much as possible by way of examples. Without departing from the inventive concept of the present invention and without creative efforts, those skilled in the art make selections and combinations of the technical features in the above embodiments, make experimental changes to the specific parameters, or form specific embodiments by making conventional substitutions of the disclosed technical means of the present invention with the existing technologies in the technical field, which should all belong to the content implicitly disclosed by the present invention.

Claims

1. An ultrasonic imaging method for a container filled with non-uniform fluid, characterized in that, Including: Step S20: Obtain the third emission position, n third sound pressure measurement positions during m emission-reception processes performed in the detection plane of the second container filled with non-uniform fluid, and the third sound pressure at the third sound pressure measurement position at the time after emission ; ;​ Step S21. Using the ultrasonic simulation model corresponding to the detection plane of the second container filled with non-uniform fluid and the data obtained in the step S21, calculate the propagation time of the ultrasonic wave emitted at the th third emission position during the m emission and reception processes to the point, and the propagation time of the ultrasonic wave from the point to the th third sound pressure measurement position, and calculate the pixel value of the P point using the full focus imaging algorithm; Traverse the points in the ultrasonic simulation model as P points to obtain the pixel values of all points in the ultrasonic simulation model, where, ; ; Step S22: Output the pixel values of all points in the ultrasonic simulation model; The method for constructing the ultrasonic simulation model includes the following steps: Step S10: Scan the first container filled with non-uniform fluid in a circular motion within the detection plane. During each emission and reception process of scanning the first container filled with non-uniform fluid in a circular motion, at the first emission position emit the ultrasonic wave, and at the time after emission, measure the first sound pressure at n first sound pressure measurement positions ; ; Step S11: Establish a scanning surface structure simulation model of the first container filled with non-uniform fluid, discretize the scanning surface structure simulation model, add the wave rules after discretization processing, and configure the sound speed of each discrete point to obtain a scanning surface sound speed field model ; Step S12: Iteratively correct the scanning plane sound speed field model using a gradient optimization algorithm , and obtain a scanning plane sound speed field model that fits the sound speed field within the scanning plane in Step S10 ; Step S13: Use the scanning plane sound velocity field model to calculate the shortest propagation duration between any two points inside it , and all the shortest propagation durations between two points in the scanning plane sound velocity field model constitute the ultrasonic simulation model; In the step S21, the full focus imaging algorithm is （5） Among them, is the pixel value at the discrete point . is the Hilbert transform, m is the number of the third emission positions, and n is the number of the third sound pressure measurement positions. is the th third sound pressure when the ultrasonic wave emitted from the th third emission position propagates to the th third sound pressure measurement position. represents the propagation time when the ultrasonic wave emitted from the th third emission position propagates to the position. represents the propagation time when the ultrasonic wave propagates from the position to the th third sound pressure measurement position.

2. The ultrasonic imaging method of a container filled with non-uniform fluid according to claim 1, characterized in that, Set In the step S12, the method of iteratively correcting the sound velocity field model of the scanning surface by using a gradient optimization algorithm comprises the following steps: Step S120, before the th iteration, simulate all the transmitting and receiving processes of step S10 in the scanning plane sound velocity field model . The scanning plane sound velocity field model outputs the second sound pressure at the corresponding position at the calculated post-transmission duration ; ; Let （1） If , the iteration is terminated, and the sound speed field model of the scanning plane is obtained ; If , the -th iteration is performed; is the maximum threshold between the second sound pressure and the first sound pressure; Step S121, at the th iteration, the gradient operator （2） Among them, is the Jacobian matrix, , is the difference between the second sound pressure and the first sound pressure output by the sound velocity field model of the scanning surface . Use the gradient optimization algorithm to find the search direction and search step, and update the sound velocity field model of the scanning plane （3） In the formula, is the iteration step size, is the search direction, ; Let , and continue to execute step S120.

3. The ultrasonic imaging method for a container filled with non-uniform fluid as claimed in claim 2, characterized in that, In the step S10, during each transmission and reception process, the ultrasonic wave is transmitted from the first transmission position in the first radiation direction, and the first transmission position, the first radiation direction, and n first sound pressure measurement positions are all set on the scanning plane; in the step S120, in the scanning plane sound velocity field model The method for simulating one transmission and reception process of the step S10 in it is: taking the relative positions of the first transmission position, n first sound pressure measurement positions and the first container filled with inhomogeneous fluid in one transmission and reception process of the step S10 as the reference basis, and respectively determining the second transmission position and n second sound pressure measurement positions according to the positions of the scanning plane structure simulation model, setting the parameters of the second ultrasonic wave simulatedly transmitted at the second transmission position, the second ultrasonic wave simulating the ultrasonic wave, the scanning plane sound velocity field model Measure the duration after transmitting the second ultrasonic wave of the second sound pressure at the second sound pressure measurement position 4. The ultrasonic imaging method for a container filled with non-uniform fluid according to claim 1, characterized in that, The container filled with the non-uniform fluid is a transformer outgoing line device. In the step S20, n > 31.

5. The ultrasonic imaging method for a container filled with non-uniform fluid according to claim 1, characterized in that, In the step S11, set the maximum grid size for discretization processing , and perform discretization processing on the simulation model of the scanning surface structure; The discrete processing fluctuation rule is （4） Among them, is the sound pressure field, are the abscissa and ordinate of discrete points in the scanning plane respectively, is the time, is the discrete point at which the sound speed is is the calculation time step, is the maximum grid size, represents 0.5 times the order of difference accuracy, is the difference coefficient.

6. The ultrasonic imaging method for a container filled with non-uniform fluid as claimed in claim 5, characterized in that, In the step S11, the maximum grid size is 1 / 8 of the minimum wavelength of the ultrasonic wave; the calculated time step satisfies , where is the minimum sound velocity of the medium in the detection plane.

7. The ultrasonic imaging method for a container filled with non-uniform fluid according to claim 6, characterized in that, In the step S13, discrete points are set within the previous sound propagation depth, corresponding to the position ; point within the subsequent sound propagation depth, corresponding to the position , then the virtual propagation duration of the two discrete points of point and point ​ In the formula, is the sound velocity at discrete points ; If points and are within the same sound propagation depth, or if points and are not within two adjacent sound propagation depths, or if points and are within two adjacent sound propagation depths, but point is within the subsequent sound propagation depth and point is within the previous sound propagation depth, then let , indicating that sound waves cannot propagate between and ; For points and not within two adjacent sound propagation depths, and the sound propagation depth to which point belongs is before the sound propagation depth to which point belongs, then use the shortest path search algorithm to calculate the shortest sound propagation duration between point and point , and update .

8. A computer-readable medium storing an ultrasonic imaging program, where the ultrasonic imaging program, when executed by a processor, implements the ultrasonic imaging method for the container filled with non-uniform fluid as described in any one of claims 1-7.

9. An ultrasonic detector, comprising a fixing frame, m ultrasonic transmitting units, n ultrasonic receiving units, and a processor, characterized in that, It further includes the computer-readable medium storing the ultrasonic imaging program as described in claim 8. The ultrasonic wave transmitting unit is used to transmit a fourth ultrasonic wave identical to the ultrasonic wave. The m ultrasonic wave transmitting units, the n ultrasonic wave receiving units and the fixing frame are fixedly connected, and the fourth ultrasonic wave of the m ultrasonic wave transmitting units is directed towards the sound pressure measurement positions of the n ultrasonic wave receiving units and are set on the same plane.

Images