A method and system for detecting the bulk density of dry material silos

Abstract

This invention discloses a method and system for detecting the bulk density of dry material silos, relating to the field of material detection technology. The method includes: after emitting an ultrasonic pulse into the material in the dry material silo, acquiring the ultrasonic echo signal reflected by the material; performing full-wave rectification processing on the ultrasonic echo signal to convert negative instantaneous amplitude values ​​in the ultrasonic echo signal to positive values; performing low-pass filtering on the rectified ultrasonic echo signal to obtain an initial energy envelope; performing noise reduction processing on the initial energy envelope to obtain a target energy envelope; calculating the target time variance of the target energy envelope; and calculating the bulk density of the material based on the target time variance. This invention can combine the energy envelope and time variance to calculate the bulk density of the material, thereby achieving bulk density detection and improving accuracy and reliability.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of material detection, and in particular to a dry material bin bulk density detection method and system. BACKGROUND

[0002] In industrial production, the bulk density of the material in the dry material bin affects the subsequent processing technology, accurate metering of the material, and the overall safety of the storage. The existing method infers the bulk density of the material by emitting an ultrasonic pulse to the material and then analyzing the received echo signal. However, in the actual industrial environment, the material in the dry material bin is not always in an ideal uniform state. The surface of the powder or granular material stored in the dry material bin can have slight hygroscopicity, adsorbing trace amounts of water molecules to form a weak surface water film, thereby generating weak capillary forces between adjacent particles. Such capillary forces cause the formation of loose, non-uniform agglomerates between particles, resulting in a large number of small pore structures of varying sizes within the material. These pores are filled with air, forming a large number of irregularly distributed acoustic impedance discontinuities with the surrounding solid particles. When the ultrasonic pulse penetrates these areas, its propagation path is significantly affected, resulting in complex scattering, reflection, and refraction phenomena, causing the ultrasonic energy to disperse, the main echo signal amplitude to attenuate, and a large number of chaotic secondary echo signals to be generated. The existing method is designed based on a relatively uniform medium model, and when dealing with a situation where the echo signal energy is severely dispersed, the detection accuracy is low and the reliability is low.

[0003] In summary, the technical problems in the related art need to be improved. SUMMARY

[0004] The main purpose of the embodiments of the present application is to provide a dry material bin bulk density detection method and system that can calculate the bulk density of the material by combining the energy envelope and the time variance to realize bulk density detection, thereby improving the accuracy and reliability.

[0005] In one aspect, the present application provides a dry material bin bulk density detection method, comprising the following steps:

[0006] After emitting an ultrasonic pulse to the material in the dry material bin, collecting the ultrasonic echo signal reflected by the material;

[0007] Full-wave rectification processing is performed on the ultrasonic echo signal, so that the negative value of the instantaneous amplitude in the ultrasonic echo signal is converted to a positive value;

[0008] Low-pass filtering is performed on the full-wave rectification processed ultrasonic echo signal to obtain an initial energy envelope;

[0009] Denoising processing is performed on the initial energy envelope to obtain a target energy envelope;

[0010] calculating a target time variance of the target energy envelope;

[0011] calculating a bulk density of the material according to the target time variance.

[0012] In another aspect, an embodiment of the present application provides a dry material bin bulk density detection system, comprising:

[0013] a signal acquisition module, configured to acquire an ultrasonic echo signal reflected by the material after emitting an ultrasonic pulse to the material in the dry material bin;

[0014] a signal rectification module, configured to perform full-wave rectification processing on the ultrasonic echo signal, so that a negative value of a transient amplitude in the ultrasonic echo signal is converted into a positive value;

[0015] an envelope extraction module, configured to perform low-pass filtering on the ultrasonic echo signal after the full-wave rectification processing, to obtain an initial energy envelope;

[0016] a denoising module, configured to perform denoising processing on the initial energy envelope, to obtain a target energy envelope;

[0017] a time variance calculation module, configured to calculate a target time variance of the target energy envelope;

[0018] a bulk density calculation module, configured to calculate a bulk density of the material according to the target time variance.

[0019] The embodiments of the present application have at least the following beneficial effects: the embodiments of the present application first acquire an ultrasonic echo signal reflected by the material, then perform full-wave rectification processing and low-pass filtering on the ultrasonic echo signal, to obtain an initial energy envelope, then perform denoising processing on the initial energy envelope, to obtain a target energy envelope, and calculate a target time variance of the target energy envelope, and finally calculate a bulk density of the material according to the target time variance, so that the bulk density of the material can be calculated in combination with the energy envelope and the time variance, to realize bulk density detection, and improve the accuracy and reliability.

[0020] Additional features and advantages of the application will be set forth in the description that follows, and in part will be apparent from the description, or can be learned by practice of the application. The objectives and other advantages of the application will be realized and attained by the structure particularly pointed out in the description and claims. BRIEF DESCRIPTION OF DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiment description will be briefly introduced.

[0022] Figure 1A flow chart of a dry material bin bulk density detection method according to an embodiment of the present application;

[0023] Figure 2 A structural schematic diagram of a dry material bin bulk density detection system according to an embodiment of the present application. DETAILED DESCRIPTION

[0024] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments.

[0025] In related art, in industrial production, the bulk density of materials in a dry material bin is a very important parameter, which directly affects the subsequent processing technology, accurate metering of materials, and overall safety of storage. In order to be able to monitor the bulk density of materials in the dry material bin in real time and without contact, a technical personnel usually adopts a method based on the characteristics of ultrasonic echo. The working principle of this method is to emit an ultrasonic pulse to the materials, and then analyze the received echo signal to infer the bulk density of the materials. However, in actual industrial environment, the materials in the dry material bin are not always in ideal uniform state, which brings challenges to traditional ultrasonic detection.

[0026] Specifically, in modern industrial production, dry material bins are widely used to store various powdery or granular materials, such as fine chemical powders, food-grade starch, or some mineral raw materials. These materials are usually required to be stored in a dry environment to maintain the stability of their physical and chemical properties. However, even under strict control of the "dry" standard, the surface of many such materials may have slight hygroscopicity. This means that during long-term storage, the surface of the material particles will inevitably adsorb trace amounts of water molecules in the air, forming an extremely weak surface water film. This water film, although difficult to detect with the naked eye, is enough to generate weak capillary forces between adjacent particles. This capillary force does not cause the material as a whole to clump or harden, but promotes the formation of loose, non-uniform agglomerates between particles in local areas. These agglomerates are not fixed and unchanging, but exhibit a state between complete free flow and complete consolidation, with a certain randomness and irregularity in their internal structure.

[0027] This agglomeration of particles by weak capillary forces results in a material that is no longer an ideal homogeneous medium, but rather a structure of numerous small pores of varying sizes. These pores are filled with air, and the pore walls are formed by the agglomerated material particles. From an acoustic propagation perspective, these air-filled pores and the surrounding solid particles form a large number of acoustically obstructive discontinuities that are irregularly distributed. Acoustic obstruction is the obstructive ability of a medium to sound wave propagation. When a sound wave enters a medium with a different acoustic obstruction from the medium it came from, part of the sound energy is reflected and part is transmitted. When an ultrasonic pulse used to detect bulk density penetrates these areas, its propagation path is significantly affected. When the ultrasonic wave encounters these interfaces with large differences in acoustic obstruction, complex scattering, reflection, and refraction phenomena occur. Part of the sound energy propagates along the expected straight path, but another part scatters in all directions, even along multiple different paths (i.e., multipath propagation) to the receiving transducer. This phenomenon causes the dispersion of the original ultrasonic energy, causing the energy of the main echo signal to be dispersed, with an amplitude decay far beyond what is expected in a homogeneous medium. At the same time, a large number of scattered and multipath propagated waves will arrive at the receiving end after the main echo, with smaller amplitudes and different delay times, forming a series of chaotic secondary echo signals.

[0028] In the face of such a complex acoustic environment, the core "echo signal flight time calculation rule" and "effective echo amplitude threshold value" of existing ultrasonic detection systems are designed based on a relatively homogeneous medium model. These systems usually expect to receive a clear, high-amplitude main echo and calculate the sound path by its arrival time to infer the material level or density. However, in the complex scattering environment caused by particle agglomeration and micro-pore structure, the energy of the main echo signal is severely dispersed, causing its amplitude to decay sharply, sometimes even below the preset effective echo amplitude threshold. More challenging is that due to scattering and multipath propagation effects, the received signal contains a large number of small-amplitude, delayed echo signals that propagate by non-direct paths. These signals are mixed with the weakened main echo, making it difficult for the system to accurately identify the starting point or peak position of the main echo that truly represents the material level. For example, an originally clear pulse echo may become indistinct after passing through such a medium, with its rising and falling edges lengthened, peak weakened, or even obscured by multiple scattering peaks, resulting in significant deviations in flight time calculations.

[0029] Furthermore, in the actual operation of the dry silo, the long-term storage of the material and the "first-in first-out" recycling mechanism will cause the material at the lower part of the silo to continuously bear the gravity extrusion from the upper layer of material. This continuous pressure effect will significantly increase the number of contact points between the particles and make the particle arrangement more compact. This physical compaction effect greatly enhances the aforementioned particle agglomeration phenomenon caused by weak capillary forces. When the particles are more closely contacted, the range and strength of the capillary force effect will increase, thereby forming "compacted layers" with different densities and significantly different physical properties at different depths of the silo, especially near the bottom and the discharge port. These compacted layers are not uniformly distributed, but may exhibit a complex structure with local high-density regions and relatively loose regions.

[0030] Due to the existence of these non-uniform compacted layers, the waveform of the ultrasonic wave will be progressively severely distorted when it propagates downward through the material layer and reflects upward, as it continuously passes through different sound impedance discontinuities. Each time it passes through a sound impedance discontinuity, it will cause partial reflection and transmission of energy, as well as slight changes in the waveform. When the ultrasonic wave continuously passes through multiple compacted layers with closely arranged and significantly different sound impedances, these effects will be cumulatively superimposed. The received echo signal will have its original pulse width significantly lengthened, and will no longer be a clear narrow pulse, but will exhibit a broadening characteristic. More importantly, the broadened waveform will contain multiple secondary peaks inside, which are reflected and superimposed by different compacted layers. The positions, relative intensities and widths of these secondary peaks reflect the trajectory of the sound wave in the complex medium. At the same time, due to the differences in attenuation and scattering characteristics of different frequency components in different density media, the frequency composition of the echo signal will also become complex, and will no longer be a simple center frequency signal. This waveform distortion caused by multiple reflections and scattering superposition has far exceeded the processing logic of existing systems that "identify a single clear reflection surface". Existing systems usually assume that the echo is from a clear interface, and the waveform characteristics are relatively simple, so it is difficult to analyze the real material level information and accurately infer the bulk density from this complex composite waveform.

[0031] Finally, under the combined action of various factors such as the slight hygroscopicity of the material itself leading to particle agglomeration characteristics, the non-uniform compaction effect formed by the long-term action of gravity in the bin, and the instantaneous structural changes during the material unloading process, and the limitations of the existing ultrasonic detection system processing logic, the detection result of the bulk density of the dry material bin becomes highly inaccurate and unreliable. Simply relying on the time of flight or the overall amplitude of the echo signal for analysis has been unable to cope with the complex echo structure caused by the compaction effect and the time-varying flow. In this harsh industrial scenario, in order to realize accurate detection of the bulk density of the dry material bin, a more in-depth morphological and structural analysis of the ultrasonic echo waveform is needed, such as analyzing the degree of pulse broadening, the number and relative energy of secondary peaks, the trend of frequency component changes, etc., to invert the true bulk density of the material and its complex internal state.

[0032] In the dry material bin, in the face of materials with slight hygroscopicity and non-uniform internal structure (such as particle agglomeration, micropore, compaction layer) formed by long-term gravity compaction, it is necessary to accurately extract and analyze the complex waveform parameters from the composite echo signal caused by ultrasonic multipath scattering and severe distortion, which can reflect the bulk density and internal structure characteristics of the material, to overcome the detection failure problem caused by signal severe attenuation, ambiguity and multipath effect of the traditional detection method based on time of flight or amplitude threshold.

[0033] Therefore, the present application first performs full-wave rectification processing on the signal after transmitting an ultrasonic pulse and collecting the echo signal, converting the negative value of the instantaneous amplitude to a positive value, so that the subsequent energy analysis is more direct and accurate, avoiding the interference of negative signals on energy calculation. Then, the rectified signal is low-pass filtered to obtain the initial energy envelope. Low-pass filtering can effectively remove high-frequency noise, thereby better extracting the trend of ultrasonic energy change over time. The initial energy envelope is then denoised to obtain the target energy envelope, to eliminate false energy peaks and noise caused by material non-uniformity or environmental interference, and to ensure that the analyzed energy envelope can truly reflect the acoustic properties of the material. After obtaining the target energy envelope, the target time variance is calculated. The target time variance, as a statistical quantity measuring the dispersion degree of the energy envelope in the time dimension, can sensitively reflect the diffusion and attenuation characteristics of the ultrasonic energy in the material. Finally, the bulk density of the material is calculated according to the target time variance. Through the pre-established mapping relationship (such as lookup table or regression model), the target time variance can be converted into a specific bulk density value of the material, thereby realizing the bulk density detection and improving the accuracy and reliability.

[0034] The embodiments of the present application will be specifically explained below in conjunction with the drawings:

[0035] Figure 1is an optional flowchart of a dry material bin bulk density detection method provided by an embodiment of the present application, Figure 1 The method in the method can include but is not limited to steps S101-S106.

[0036] Step S101, after emitting an ultrasonic pulse to the material in the dry material bin, collecting the ultrasonic echo signal reflected by the material;

[0037] Step S102, full-wave rectification processing is performed on the ultrasonic echo signal, so that the negative value of the instantaneous amplitude in the ultrasonic echo signal is converted to a positive value;

[0038] Step S103, low-pass filtering is performed on the ultrasonic echo signal after full-wave rectification processing, to obtain an initial energy envelope line;

[0039] Step S104, denoising processing is performed on the initial energy envelope line to obtain a target energy envelope line;

[0040] Step S105, calculating the target time variance of the target energy envelope line;

[0041] Step S106, calculating the material bulk density according to the target time variance.

[0042] The steps S101-S106 shown in the embodiment of the present application can combine the energy envelope line and the time variance to calculate the material bulk density, so as to realize bulk density detection and improve the accuracy and reliability.

[0043] In some embodiments, steps S101-S106 can first collect the ultrasonic echo signal reflected by the material after emitting an ultrasonic pulse to the material in the dry material bin. For example, an ultrasonic sensor array can be used for signal collection. For example, a transmitter and multiple receivers can be provided, the transmitter periodically emits ultrasonic pulses to the material in the dry material bin, and the receivers synchronously receive the echo signals reflected by the material. These receivers can be uniformly distributed on the side wall or top of the dry material bin to obtain echo information at different angles and depths. In another embodiment, a single ultrasonic sensor with transmitting and receiving functions can be used, which is responsible for both emitting ultrasonic pulses and receiving reflected ultrasonic echo signals. After receiving the echo signals, the signals are digitized and stored for subsequent processing. It can be understood that the ultrasonic pulse refers to a short-time, high-frequency acoustic signal that is reflected when it encounters different acoustic impedance interfaces in the medium. The ultrasonic echo signal refers to the signal received by the receiver after the ultrasonic pulse propagates in the material and is reflected by the material.

[0044] The collected ultrasonic echo signal usually contains positive and negative instantaneous amplitude values. In order to facilitate energy analysis, the ultrasonic echo signal can be subjected to full-wave rectification processing, so that the negative value of the instantaneous amplitude in the ultrasonic echo signal is converted to a positive value. An absolute value function can be used to process the signal, and the absolute value of the instantaneous amplitude value of each sampling point is taken. For example, if the instantaneous amplitude of a certain sampling point is -0.5V, it becomes 0.5V after full-wave rectification. In another embodiment, a bridge rectifier circuit is used to convert an alternating current signal into a unidirectional pulsating direct current signal, and then a filter is used to obtain a smooth energy envelope.

[0045] The ultrasonic echo signal subjected to full-wave rectification processing is then subjected to low-pass filtering to obtain an initial energy envelope. The purpose of low-pass filtering is to remove high-frequency noise and burrs in the signal, thereby better revealing the energy trend of the signal. For example, a moving average filter can be used to smooth the signal by calculating the average value of the signal within a certain time window. In another embodiment, a digital filter such as a Butterworth filter or a Chebyshev filter can be used, which can effectively filter out high-frequency components and retain low-frequency energy envelope information according to the pre-set cutoff frequency and order. At the same time, the initial energy envelope is subjected to denoising processing to obtain a target energy envelope. The purpose of denoising processing is to eliminate false energy peaks or noise in the initial energy envelope caused by environmental interference or material non-uniformity, thereby obtaining a more accurate energy distribution curve. For example, a threshold denoising method can be used, in which an energy threshold is set, and energy envelope values below the threshold are considered as noise and are set to zero or attenuated. The energy threshold can be estimated based on the noise level. A segment containing only noise can be extracted from the signal, and the average energy envelope value of the segment containing only noise is calculated. The average energy envelope value is multiplied by a fixed multiple to obtain the energy threshold. The fixed multiple can be determined by experiment or expert experience, aiming to effectively suppress noise while not affecting low-energy useful signals. For example, environmental noise data can be collected and statistically analyzed to calculate the mean and standard deviation, and the maximum value of the environmental noise data plus the standard deviation divided by the mean value can be used to obtain the fixed multiple. In another embodiment, wavelet transform denoising can be used, in which the signal is decomposed into wavelet coefficients at different scales, and the wavelet coefficients corresponding to noise are subjected to threshold processing or shrinkage, and the signal is reconstructed to achieve denoising. The wavelet coefficients are obtained by wavelet transform of the original signal. Wavelet transform is a mathematical tool that decomposes a signal into different frequency (or scale) components. A suitable wavelet basis function can be selected, such as Haar wavelet, Daubechies wavelet, etc. In one iteration, the signal is passed through a low-pass filter and down-sampled to obtain approximation coefficients, and the signal is passed through a high-pass filter and down-sampled to obtain detail coefficients. The calculated approximation coefficients are used as input for the next iteration, and the approximation coefficients and detail coefficients of the next iteration are obtained. These detail coefficients and approximation coefficients are collectively referred to as wavelet coefficients.

[0046] The target time variance of the target energy envelope is calculated. The target time variance is a statistical quantity that measures the dispersion degree of the energy envelope in the time axis, which can reflect the energy diffusion and attenuation characteristics of the ultrasonic wave in the material. For example, the target energy envelope can be regarded as a probability density function, and the second-order central moment, i.e., the time variance, is calculated. Specifically, the centroid time of the energy envelope can be calculated, the square of the difference between the time of each sampling point and the centroid time is calculated, and the energy value of the corresponding sampling point is multiplied, the sum of all products is normalized to obtain the target time variance. It can be understood that the target time variance is a parameter for measuring the dispersion degree of the energy distribution of the target energy envelope in the time dimension, which is related to the bulk density of the material.

[0047] Finally, the bulk density of the material is calculated according to the target time variance. There is a certain mapping relationship between the target time variance and the bulk density of the material. The target time variance can be associated with the bulk density of the material through pre-established experimental data or theoretical models. For example, a lookup table can be established, which contains the bulk density of the material corresponding to different target time variance values. In another embodiment, a mathematical model can be established through regression analysis, such as a linear regression or a nonlinear regression model, taking the target time variance as the input and outputting the bulk density of the material.

[0048] Through the above technical solutions, the energy envelope is first stably extracted through full-wave rectification and low-pass filtering, avoiding excessive dependence on a single main echo. Then, the initial energy envelope is denoised to effectively eliminate interference signals and obtain a purer target energy envelope. The target time variance is calculated, which can comprehensively reflect the overall propagation characteristics of the ultrasonic wave in the material, including scattering, attenuation, and the dispersion degree of energy distribution, thereby more robustly representing the bulk density of the material. For example, in the scenario where material particles agglomerate to form a large number of micro-pores, the traditional method may not be able to identify the main echo due to its low amplitude, or the time of flight may be calculated incorrectly due to multipath effects. However, the present embodiment can accurately infer the bulk density through the overall diffusion degree of energy even if the main echo is not obvious, thereby avoiding strict dependence on specific echo characteristics and significantly improving detection accuracy and reliability in complex and non-uniform material environments, providing reliable technical support for accurate management of dry material bins.

[0049] In some embodiments, in step S104, the initial energy envelope is denoised to obtain the target energy envelope, which can include but is not limited to the following steps:

[0050] The initial energy envelope is segmented to obtain a plurality of candidate energy regions.

[0051] determine an energy threshold according to the overall average energy and the local energy fluctuation degree of the initial energy envelope;

[0052] select an energy region from the plurality of candidate energy regions as a to-be-identified energy region;

[0053] if the energy envelope value in the to-be-identified energy region is greater than the energy threshold, determine whether the duration of the to-be-identified energy region is greater than a first preset duration threshold;

[0054] if the duration of the to-be-identified energy region is greater than the first preset duration threshold, determine whether the energy concentration degree of the to-be-identified energy region is less than a preset concentration threshold;

[0055] if the energy concentration degree of the to-be-identified energy region is less than the preset concentration threshold, take the to-be-identified energy region as a wide secondary energy region;

[0056] perform energy weighting attenuation processing on the wide secondary energy region in the initial energy envelope to obtain a target energy envelope.

[0057] In some embodiments, the wide secondary energy region may not be caused by the real reflection of the material, but by a pseudo-signal caused by factors such as the internal structure of the dry material bin, multipath effect or environmental interference. If not effectively identified and processed, these pseudo-signals will confuse the real energy envelope, causing distortion of the target time variance calculated subsequently, and thus affecting the accuracy of the material bulk density detection. Therefore, the initial energy envelope can be processed in segments first to obtain a plurality of candidate energy regions. For example, the continuous initial energy envelope can be divided into a plurality of independent or overlapping energy segments according to a preset time interval or energy change characteristic, and each energy segment is a candidate energy region. The purpose is to decompose the complex energy envelope into local units that are easier to analyze and process.

[0058] Then, an energy threshold is determined according to the overall average energy and the local energy fluctuation degree of the initial energy envelope. The threshold can be dynamically or statically set to distinguish between valid signals and potential interference. For example, the root mean square energy or average energy of the initial energy envelope can be calculated, and a self-adaptive threshold can be determined in combination with the standard deviation. The purpose is to provide a basis for subsequent interference identification. Then, a to-be-identified energy region is selected from the plurality of candidate energy regions. Each segmented energy region can be checked in time sequence or other preset rules for further analysis. Each candidate energy region can be analyzed subsequently.

[0059] If the energy envelope value in the to-be-identified energy region is greater than the energy threshold value, it is judged whether the duration of the to-be-identified energy region is greater than a first preset duration threshold value. The purpose is to preliminarily screen out signals with certain energy intensity and duration, and exclude transient noise. The first preset duration threshold value can be set according to the actual application scene and experience, for example, it can be set to a certain proportion of the time required for ultrasonic waves to propagate a typical reflection path in the dry material bin. If the duration of the to-be-identified energy region is greater than the first preset duration threshold value, it is judged whether the energy concentration degree of the to-be-identified energy region is less than a preset concentration threshold value. The energy concentration degree can be obtained by calculating the variance or entropy value of the energy in the region. The smaller the variance or the lower the entropy value, the more concentrated the energy. The preset concentration threshold value is used to distinguish between primary reflection signals and wide secondary interference signals, and the purpose is to identify those non-typical reflection signals with dispersed energy distribution and long duration. The preset concentration threshold value can be determined by empirical statistical method. A large number of energy region data samples representing different conditions (including concentrated and not concentrated) can be collected first. The energy concentration values of these data samples are calculated respectively, and the distribution characteristics of the energy concentration values of different definitions are analyzed. The target energy region and the background noise will be in two different ranges respectively, and a critical point that can effectively distinguish the two different ranges is selected as the preset concentration threshold value. If the energy concentration degree of the to-be-identified energy region is less than the preset concentration threshold value, the to-be-identified energy region is regarded as a wide secondary energy region. The wide secondary energy region is usually characterized by non-concentrated energy distribution, long duration, and possibly high amplitude, but it is not a primary reflection wave.

[0060] The wide secondary energy region in the initial energy envelope line is further subjected to energy weighted attenuation processing to obtain a target energy envelope line. Exemplarily, the energy value of the identified wide secondary energy region can be multiplied by an attenuation coefficient less than 1, or a nonlinear attenuation function can be used to reduce its influence on the overall energy envelope line. The attenuation coefficient can be dynamically adjusted according to the amplitude, duration or concentration degree of the region, and the purpose is to effectively suppress interference without completely deleting the signal, so as to obtain a more pure and accurate target energy envelope line.

[0061] The embodiment can effectively identify and suppress the wide secondary energy region by fine segmentation processing and multi-dimensional feature judgment on the initial energy envelope. Specifically, first, the complex energy envelope is decomposed into manageable candidate regions through segmentation processing, laying the foundation for local analysis. Second, the energy threshold is determined in combination with the overall average energy and the local energy fluctuation degree, so that the threshold can adapt to different signal environments. Subsequently, the energy envelope value, duration and energy concentration degree of the candidate region are judged step by step, which can accurately distinguish the wide secondary energy region with certain intensity, long duration but uneven energy distribution. These regions are often caused by multi-path effect, structural reflection or environmental noise and other non-main reflection sources. Finally, the energy weighting attenuation processing is performed on the identified wide secondary energy region, rather than simply deleting it, which not only preserves the integrity of the original signal, but also significantly reduces the influence of interference on the target energy envelope, thereby ensuring the accuracy of subsequent time variance calculation.

[0062] To more clearly illustrate the technical solutions, specific examples are used in the following. Assuming that after emitting an ultrasonic pulse in the dry material bin, the collected ultrasonic echo signal is subjected to full-wave rectification and low-pass filtering, an initial energy envelope is obtained. After the main peak, there is a secondary energy region with relatively high amplitude, long duration but uneven energy distribution. First, the initial energy envelope is segmented, for example, divided into a candidate energy region every 100 sampling points. Then, the overall average energy and local energy fluctuation degree of the initial energy envelope are calculated, and an energy threshold is determined therefrom, for example, set to 1.5 times the overall average energy. Then, the system selects a region from the candidate energy regions as a to-be-identified energy region for analysis. For example, when analyzing the secondary energy region after the main peak, it is found that the energy envelope value is greater than the set energy threshold. Further, the system judges the duration of the to-be-identified energy region. If the duration is greater than a first preset duration threshold (for example, 50 milliseconds), the judgment continues. Subsequently, the system calculates the energy concentration degree of the to-be-identified energy region, for example, by calculating its energy variance. If the energy variance is less than a preset concentration threshold (for example, 0.05), it indicates that the energy distribution of the region is relatively dispersed. Based on the above judgment, the to-be-identified energy region is confirmed as a wide secondary energy region. Finally, the energy weighting attenuation processing is performed on the wide secondary energy region. For example, the energy value can be multiplied by an attenuation coefficient 0.3, thereby reducing its influence on the overall energy envelope. In this way, the original initial energy envelope is modified to obtain a more accurate target energy envelope that removes the main interference, providing a reliable data basis for subsequent bulk density calculation.

[0063] Through the above technical solution, this embodiment, by introducing segmented processing, multi-condition judgment (including energy threshold, duration threshold, and energy concentration threshold), and energy weighted attenuation, can accurately identify and weaken wide-range secondary energy regions that may lead to measurement errors. As a result, the obtained target energy envelope has higher purity and accuracy, significantly improving the reliability of subsequent target time variance calculation. This makes the detection results of material bulk density in the dry material silo more accurate and stable, reduces the false positive rate, and improves the overall performance of the system.

[0064] In some embodiments, after denoising the initial energy envelope to obtain the target energy envelope, the following steps may be included, but are not limited to:

[0065] Identify the secondary energy peak in the target energy envelope. The duration of the secondary energy peak is less than a preset time window, and the amplitude of the secondary energy peak is greater than a preset amplitude threshold.

[0066] If the secondary energy peak is located in the front region of the target energy envelope, the secondary energy peak is suppressed and the target energy envelope is updated.

[0067] In some embodiments, the target energy envelope obtained after denoising may still contain some secondary energy peaks caused by non-primary reflections or transient disturbances. These secondary energy peaks, especially when they appear in the leading region of the energy envelope, may interfere with the accurate identification of the primary energy peak and subsequent calculation of the target time variance, thereby affecting the accuracy of material bulk density detection.

[0068] To address this, secondary energy peaks within the target energy envelope can be identified first. Specific signal processing algorithms or preset judgment rules can be used to detect and distinguish non-primary energy peaks with specific energy characteristics from the target energy envelope. The duration of the secondary energy peaks is shorter than a preset time window, designed to differentiate short-duration transient interference signals from longer-duration primary energy reflection signals. The preset time window can be set according to the actual application scenario and the characteristics of the ultrasonic signal. Simultaneously, the amplitude of the secondary energy peaks is greater than a preset amplitude threshold, used to filter out secondary peaks with sufficient energy intensity to affect subsequent calculations, thereby avoiding over-processing of weak noise. The preset amplitude threshold can also be determined based on experience or experimental data.

[0069] Then, it is determined whether the occurrence time of the secondary energy peak falls within a predetermined time period at the beginning of the entire energy envelope. For example, this initial region can be defined as a certain time proportion range between the ultrasonic pulse emission time and the arrival time of the main energy peak. Secondary energy peaks located in the initial region are usually more likely to be caused by reflections or near-field effects near the transmitter, and have a greater impact on the early energy distribution of the ultrasonic signal. If the secondary energy peak is located in the initial region of the target energy envelope, the secondary energy peak is suppressed, and the target energy envelope is updated. For example, the energy envelope value in the time region where the secondary energy peak is located can be set to zero to completely eliminate its influence; or, its amplitude can be reduced by applying a weighted attenuation function, a smoothing filter, etc., so that its influence on the overall energy envelope is reduced to an acceptable level. After suppression, the target energy envelope will be updated to form a purer and more accurate energy envelope for subsequent target time variance calculation.

[0070] This embodiment effectively solves the calculation error problem caused by these interference peaks that may exist in traditional methods by further identifying and suppressing secondary energy peaks, especially those located in the front region, after obtaining the target energy envelope. Secondary energy peaks, especially those with short durations but large amplitudes, may be mistakenly identified as part of the valid signal or interfere with the accurate identification of the main energy peak if not processed, causing the calculated target time variance to deviate from the true value. By setting a preset time window and a preset amplitude threshold, these secondary energy peaks with interference characteristics can be accurately screened out. When these secondary energy peaks are located in the front region of the target energy envelope, their influence on the initial arrival time or early energy distribution of the ultrasonic signal is particularly significant. Suppressing them ensures that the subsequent target time variance calculation is based on a more realistic and less interference-free energy envelope, thereby improving the robustness and accuracy of the entire packing density detection process.

[0071] To illustrate this technical solution more clearly, a specific example is used below. Suppose that after denoising the initial energy envelope, a secondary energy peak with a duration of 5 microseconds and an amplitude of 0.2V appears before the main energy peak in the target energy envelope. At this point, the preset time window is set to 10 microseconds, and the preset amplitude threshold is set to 0.1V. Because the duration of this secondary energy peak (5 microseconds) is less than the preset time window (10 microseconds), and its amplitude (0.2V) is greater than the preset amplitude threshold (0.1V), and it is located in the leading region of the target energy envelope (e.g., within the first 20% of the arrival time of the main energy peak), this secondary energy peak is identified as interference that needs to be suppressed. The system then suppresses the region containing this secondary energy peak, for example, by setting the energy envelope value of that region to zero, or by smoothing it with a low-pass filter to significantly reduce its amplitude. After suppression processing, the target energy envelope is updated, and the influence of secondary energy peaks in the front region is effectively eliminated, thus providing a clearer and more accurate energy envelope for subsequent target time variance calculation.

[0072] Through the above technical solution, this embodiment can effectively eliminate the influence of secondary energy peaks caused by non-primary reflections or transient interference in the target energy envelope, especially by specifically addressing interference in the front region. This significantly improves the purity and reliability of the target energy envelope, enabling a more accurate reflection of the material's true energy distribution characteristics when subsequently calculating the target time variance. Consequently, it avoids errors in bulk density calculation caused by secondary energy peak interference, improves the accuracy and stability of bulk density detection in dry material silos, and provides more reliable data support for the precise management of dry material silos.

[0073] In some embodiments, calculating the target time variance of the target energy envelope in step S105 may include, but is not limited to, the following steps:

[0074] Identify the energy concentration region within the target energy envelope;

[0075] Calculate the local time variance of the energy concentration region;

[0076] Calculate the corrected time variance based on the local time variance and energy weight;

[0077] Calculate the target time variance based on the corrected time variance.

[0078] In some embodiments, since the energy distribution of the target energy envelope may be uneven, directly calculating the time variance over the entire envelope may be affected by non-critical energy regions, leading to a decrease in the accuracy of the calculation results. Therefore, energy concentration regions within the target energy envelope can be identified first. For example, by analyzing the instantaneous energy values ​​of the target energy envelope, continuous time periods where the energy is significantly higher than the background or average level can be determined. For instance, peak detection algorithms can be used to identify the primary energy peak and its nearby secondary peaks, or an energy threshold can be set, defining regions where the energy envelope value is consistently higher than this threshold as energy concentration regions. The aim is to separate the energy portion that contributes the most to the material's bulk density from the overall envelope, thereby reducing interference from non-critical regions.

[0079] Then, the local temporal variance of the energy concentration region is calculated, allowing for the independent calculation of the temporal dispersion of the energy distribution within each identified energy concentration region. Specifically, the weighted squared difference between the time of each sampling point within the energy concentration region and the time of the energy centroid of that region can be calculated, where the weights can be determined by the energy envelope value of each sampling point. The purpose is to accurately quantify the degree of energy dispersion on the time axis within each energy concentration region, providing basic data for subsequent corrections.

[0080] Then, based on the local time variance and energy weights, the corrected time variance is calculated. The local time variances of each energy concentration region can be weighted and averaged or summed according to their respective energy weights. For example, the energy weight can be determined based on the total energy, peak amplitude, or duration of each energy concentration region; regions with higher energy have greater weights. The purpose is to comprehensively consider the contribution of different energy concentration regions to the overall time variance, ensuring that regions with larger energy contributions dominate the final corrected time variance, thereby improving the representativeness of the calculation results.

[0081] Finally, the target time variance is calculated based on the corrected time variance. For example, the time variance corrected for energy weights can be used as the final target time variance. This ensures that the obtained target time variance can more accurately and robustly reflect the bulk density characteristics of the material, providing reliable input for subsequent bulk density calculations.

[0082] To illustrate this technical solution more clearly, a specific example is used below. Assume the target energy envelope is sampled as a series of discrete time points and corresponding energy values. First, by setting a dynamic threshold (e.g., based on a certain percentage of the average energy of the envelope), all consecutive time periods with energy values ​​higher than this threshold are identified; these time periods are considered energy concentration regions. For each identified energy concentration region, the time-weighted average of all sampling points within it is calculated as the energy centroid time of that region. Furthermore, the weighted sum of squares of the time of each sampling point relative to this centroid time is calculated to obtain the local time variance of that region. Subsequently, the total energy of each energy concentration region is calculated, and the proportion of this total energy to the total energy of all energy concentration regions is used as the energy weight of that region. Finally, each local time variance is multiplied by its corresponding energy weight and summed to obtain the corrected time variance, which is used as the final target time variance.

[0083] Through the above technical solution, this embodiment can effectively avoid interference from non-primary energy regions or noise in the target energy envelope on the time variance calculation, enabling the calculated target time variance to more accurately reflect the true energy distribution characteristics of the material. Especially in cases of uneven energy distribution or multiple peaks, by performing local calculations on energy-concentrated regions and combining them with energy weights for correction, the accuracy and anti-interference capability of bulk density detection are significantly improved, thus providing more reliable input parameters for subsequent bulk density calculations.

[0084] In some embodiments, calculating the target time variance of the target energy envelope in step S105 may include, but is not limited to, the following steps:

[0085] Identify the trailing region after the main energy peak in the target energy envelope. The amplitude of the trailing region is smaller than that of the main energy peak, and the duration of the trailing region is greater than a second preset duration threshold.

[0086] Calculate the product of amplitude and time based on the amplitude and duration of the trailing region;

[0087] Calculate the correction weight for the trailing region based on the product of amplitude and time;

[0088] The target time variance is calculated based on the energy and corrected weights of the trailing region.

[0089] In some embodiments, due to complex effects such as scattering and absorption when ultrasonic signals propagate through materials, a tail region containing rich material characteristic information often forms after the main energy peak. If this tail region is not fully utilized or is not properly processed, the calculated target time variance may not accurately reflect the true bulk density of the material, especially when the particle morphology, moisture content, and other characteristics of the material change, the detection accuracy will be affected.

[0090] To address this, we can first identify the trailing region following the main energy peak in the target energy envelope. The amplitude of the trailing region is smaller than that of the main energy peak, and its duration is greater than a second preset duration threshold. This trailing region typically refers to the subsequent portion of the ultrasonic signal energy that slowly decays after the main energy arrives and attenuates, due to multiple scattering, reflection, and energy dissipation within the material. When identifying this trailing region, its amplitude can be set to be smaller than that of the main energy peak to distinguish it from the main peak. Simultaneously, to ensure the identified trailing region is sufficiently representative, its duration should be greater than the second preset duration threshold. This second preset duration threshold can be empirically set or determined experimentally based on the actual application scenario and material characteristics.

[0091] Then, based on the amplitude and duration of the trailing region, the amplitude-time product is calculated. The amplitude-time product is a comprehensive representation of the energy information contained in the trailing region in both time and amplitude dimensions, providing a more complete picture of the trailing region's characteristics. For example, the amplitude-time product can be obtained by integrating the energy envelope value within the trailing region, or by multiplying the average amplitude of the trailing region by its duration.

[0092] Next, based on the product of amplitude and time, a corrected weight for the tail region is calculated to quantify its contribution to the overall time variance calculation. A larger product of amplitude and time indicates richer energy information in the tail region, potentially leading to a stronger indication of packing density; therefore, a larger corrected weight can be assigned. The corrected weight can be calculated using a linear or nonlinear mapping function, converting the amplitude-time product into a weight value between 0 and 1.

[0093] Finally, based on the energy of the tail region and the corrected weights, the target time variance is calculated. The corrected weights can be used for weighted processing to more accurately reflect its impact on the material's bulk density. For example, the weighted energy of the tail region can be combined with the energy of the main energy peak region to jointly calculate the target time variance, or the weighted energy of the tail region can be directly used as part of the target time variance.

[0094] To illustrate this technical solution more clearly, a specific example is used below. Suppose that in an ultrasonic test, the acquired target energy envelope shows a distinct main energy peak, followed by a prolonged tail region. First, the system automatically identifies this tail region, for example, by setting an amplitude threshold (e.g., 20% of the main peak amplitude) and a duration threshold (e.g., 50 microseconds). If the amplitude of the tail region is below 20% of the main peak amplitude and its duration exceeds 50 microseconds, it is identified as a tail region. Next, the tail region is analyzed, and its average amplitude is calculated. The duration is Then, calculate the product of amplitude and time. For example, if It is 0.1V. If it is 100 microseconds, then It is 10.

[0095] Subsequently, according to the preset mapping relationship, the amplitude is multiplied by time. Convert to corrected weights For example, a linear mapping function can be defined: ,in and These are preset parameters. If... The larger, The larger the value, the greater the contribution of the tail region to the time variance. Preset parameters can be obtained by fitting experimental data, allowing for the measurement and collection of a series of data under different operating conditions. Value and corresponding The values ​​are used to form a set of data points. Statistical methods, such as least squares, are used to perform a linear fit on this set of data points to obtain the value that best represents the linear trend of the data points. Value and Value. Finally, the energy of the trailing region. With corrected weights Combined, the target time variance is calculated. For example, the weighted energy of the trailing region ( * Energy in the main peak region A comprehensive calculation is performed to obtain the final target time variance. This weighted processing ensures that the key information about material bulk density carried by the tail region is fully and accurately utilized, thereby improving the overall detection accuracy.

[0096] Through the above technical solution, this embodiment can fully utilize the information about the microstructure and scattering characteristics of the material contained in the ultrasonic signal by identifying and weighting the tail region after the main energy peak. This allows the calculated target time variance to more accurately characterize the bulk density of the material and effectively reduce the detection error caused by changes in material particle shape, moisture content and other factors. It is especially suitable for industrial application scenarios with high detection accuracy requirements.

[0097] In some embodiments, calculating the target time variance of the target energy envelope in step S105 may include, but is not limited to, the following steps:

[0098] Multiple sampling points are selected from the target energy envelope;

[0099] Calculate the centroid moment of the target energy envelope based on the energy envelope value corresponding to each sampling point and the total number of sampling points;

[0100] Based on the centroid time and the point weight corresponding to each sampling point, the weighted sum of squares is calculated as the target time variance.

[0101] In some embodiments, using only simple statistical methods to calculate the time variance may not fully capture the fine distribution characteristics of the energy envelope, especially when the signal has fluctuations or noise. This could affect the accuracy and stability of the calculation results, and consequently the accuracy of the inference of the material's bulk density. Therefore, multiple sampling points can be selected from the target energy envelope. For example, a series of discrete time points can be selected according to a certain sampling frequency or interval. These sampling points are used for subsequent calculations to characterize the shape and distribution of the entire energy envelope. The energy envelope value corresponding to each sampling point refers to the amplitude value of the target energy envelope at that sampling time, reflecting the energy intensity of the ultrasonic echo signal at that moment. The total number of sampling points refers to the total number of discrete points selected on the target energy envelope.

[0102] Then, based on the energy envelope value corresponding to each sampling point and the total number of sampling points, the centroid time of the target energy envelope is calculated. This can be understood as calculating the "centroid" position of the energy envelope on the time axis. It can be obtained by weighted averaging the energy envelope value of each sampling point and its corresponding time; that is, the centroid time equals the sum of the products of the time and energy envelope value of all sampling points divided by the sum of the energy envelope values ​​of all sampling points. The centroid time reflects the central position of the overall energy distribution of the energy envelope.

[0103] Then, based on the centroid time and the point weight corresponding to each sampling point, a weighted sum of squares is calculated as the target time variance. The difference between the time of each sampling point and the centroid time can be squared, multiplied by the point weight of that sampling point, and all products summed. This allows for a more accurate measurement of the dispersion of the energy envelope relative to its centroid time, i.e., the "width" or "dispersion" of the energy distribution. The point weight of each sampling point can be determined based on its importance in the energy envelope or its energy contribution. For example, the energy envelope value of each sampling point can be simply used as its point weight, or other functional forms can be used to assign weights to highlight high-energy regions or provide more refined consideration of specific regions.

[0104] To illustrate this technical solution more clearly, a specific example is used below. Suppose that at a certain moment, the target energy envelope obtained through denoising is digitized into a series of discrete sampling points; for example, N sampling points are selected at equal intervals along the time axis. , , ..., Each sampling point corresponds to an energy envelope value. , , ..., First, the moment of the center of mass can be calculated using the following formula: In the formula, This indicates summing over all sampling points. For the moment of the center of mass, For the i-th sampling point, Let be the energy envelope value corresponding to the i-th sampling point. Next, the point weights for each sampling point can be defined. For example, the energy envelope value can be directly expressed. As point weights, i.e. = Then, the weighted sum of squares is calculated as the target time variance: In the formula, For the target time variance, The sampling points are weighted. In this way, sampling points with higher energy contribute more to the time variance, making the calculation results more reflective of the distribution characteristics of the energy matrix and effectively avoiding the excessive influence of low-energy regions or noise on the overall variance calculation. For example, when the target energy envelope exhibits a narrow and concentrated energy distribution, the calculated target time variance will be smaller; while when the energy distribution is wider, the target time variance will be larger, thus accurately reflecting the changes in material bulk density.

[0105] Through the above technical solution, this embodiment introduces the concepts of centroid time and weighted sum of squares, enabling a more detailed characterization of the dispersion of ultrasonic echo signal energy distribution. This not only improves the anti-interference capability of time variance calculation and reduces the influence of noise and secondary energy peaks on the results, but also allows the calculated time variance to more accurately reflect the true changes in material bulk density, thereby significantly improving the accuracy and reliability of dry silo bulk density detection.

[0106] In some embodiments, calculating the target time variance of the target energy envelope in step S105 may include, but is not limited to, the following steps:

[0107] Obtain the amplitude and arrival time of the main energy peak in the target energy envelope;

[0108] If the amplitude of the main energy peak is less than the preset reference amplitude and the arrival time is different from the preset reference time, then adjust the ultrasonic wave emission direction or receiving gain and update the target energy envelope.

[0109] Calculate the target time variance based on the updated target energy envelope.

[0110] In some embodiments, the quality of the ultrasonic signal directly affects the accuracy of the target energy envelope, and consequently the accuracy of the target time variance calculation. For example, when an ultrasonic signal propagates within a dry material silo, it may be affected by material characteristics, environmental factors, or sensor status, resulting in insufficient received echo signal strength or abnormal propagation time. This can cause the amplitude of the main energy peak in the target energy envelope to be too low or the arrival time to deviate from expectations, thereby introducing calculation errors. This may lead to inaccurate detection results of material bulk density, affecting production process control and material management.

[0111] To achieve this, the amplitude and arrival time of the main energy peak in the target energy envelope can be obtained first. The arrival time of the main energy peak refers to the start or peak time of the main energy peak. It can be extracted from the target energy envelope using peak detection algorithms or energy centroid algorithms. If the amplitude of the main energy peak is less than the preset reference amplitude and the arrival time is different from the preset reference time, it indicates that the current ultrasonic signal quality may be poor or the propagation path may be abnormal. In this case, the ultrasonic emission direction or receiving gain needs to be adjusted. Adjusting the ultrasonic emission direction aims to optimize the propagation path of the ultrasonic wave in the material and reduce scattering and attenuation, for example, through mechanical adjustment or electronic beam control. Adjusting the receiving gain aims to compensate for signal attenuation and ensure that the received echo signal has a suitable amplitude range, avoiding a weak or saturated signal. After adjustment, the ultrasonic pulse needs to be re-emitted and the echo signal needs to be acquired. After the above full-wave rectification, low-pass filtering, and noise reduction processing, the updated target energy envelope is obtained. The preset reference amplitude refers to the minimum amplitude value that the main energy peak should reach under ideal or standard operating conditions, used to ensure that the signal has sufficient strength for reliable analysis. The preset reference time refers to the expected time when the main energy peak should arrive under standard conditions, used to determine whether there are any abnormalities in the signal propagation path or medium. These reference values ​​are usually preset through experimental calibration or based on the structure and material characteristics of the dry material silo. Then, the target time variance is calculated based on the updated target energy envelope.

[0112] To illustrate this technical solution more clearly, a specific example is used below. Suppose that during material bulk density detection in a dry material silo, the unevenness of the material surface or slight misalignment of the ultrasonic sensor results in a weak ultrasonic echo signal. After full-wave rectification, low-pass filtering, and noise reduction, the amplitude of the main energy peak in the obtained target energy envelope is detected as lower than a preset reference amplitude. For example, the preset reference amplitude is 50mV, while the actual detected amplitude is only 30mV. Simultaneously, the arrival time of this main energy peak is also 100 microseconds later than the preset reference time. The system will determine that the current signal quality does not meet the requirements. At this point, the control system can automatically adjust the direction of the ultrasonic transmitter to more accurately align it with the material surface, or increase the gain of the ultrasonic receiver. For example, increasing the receiver gain from the default 20dB to 30dB. After adjustment, the system will re-emit the ultrasonic pulse and acquire the new echo signal, process it again, and obtain an updated target energy envelope. At this point, the amplitude of the main energy peak in the updated target energy envelope may reach 60mV, and the arrival time will also return to near the preset reference time. Based on this optimized target energy envelope, the target time variance is then calculated to obtain a more accurate material bulk density. In this way, under non-ideal detection conditions, this embodiment can ensure the accuracy and reliability of the detection results.

[0113] Through the above technical solution, this embodiment monitors the amplitude and arrival time of the main energy peak in real time and judges based on preset thresholds. Once poor signal quality is detected, the ultrasonic emission direction or receiving gain can be adjusted in a timely manner to obtain a higher quality and more representative target energy envelope. This adaptive signal optimization mechanism significantly improves the accuracy of target time variance calculation, thereby ensuring the stability and reliability of material bulk density detection results and avoiding misjudgments or accuracy reduction caused by signal quality issues. This makes the method more robust and applicable in practical industrial applications.

[0114] In some embodiments, in step S106, calculating the material bulk density based on the target time variance may include, but is not limited to, the following steps:

[0115] Obtain information on the type, moisture content, and particle morphology of the material;

[0116] Based on the type information, moisture content information and particle morphology information, the corresponding inference parameters are selected from the preset inference parameter set. The preset inference parameter set includes multiple linear mapping parameters or nonlinear mapping functions for different combinations of material types, moisture content and particle morphology.

[0117] Based on the inferred parameters, the target time variance is converted into the material bulk density.

[0118] In some embodiments, factors such as the type of material, moisture content, and particle morphology can significantly affect the propagation characteristics of ultrasound in the material, making it difficult for a single conversion relationship to accurately reflect the bulk density under all operating conditions. Directly converting the target time variance to bulk density without considering these material characteristics may introduce significant errors, affecting the accuracy and reliability of the detection.

[0119] To achieve this, information on the material type, moisture content, and particle morphology can be obtained first. Material type refers to the specific type of material stored in the dry material silo, such as pulverized coal, cement, slag, and grains. Different types of materials have significantly different physicochemical properties, affecting the attenuation and propagation speed of ultrasonic waves differently. Moisture content refers to the amount of water in the material. The presence of water alters the acoustic impedance and internal structure of the material, thus affecting the characteristics of the ultrasonic echo signal. Particle morphology information includes the size, shape, and surface roughness of the material particles. These microscopic characteristics also affect the scattering and absorption of ultrasonic waves within the material. This information can be obtained through sensors (such as humidity sensors or visual recognition systems) or through manual input or database queries.

[0120] Then, based on the type, moisture content, and particle morphology information, corresponding inference parameters are selected from a preset inference parameter set. This preset inference parameter set is a pre-established set of parameters or functions used to convert the target time variance into material bulk density. This parameter set is formed based on extensive experimental data and accumulated experience, aiming to establish a mapping relationship between the target time variance and bulk density. Specifically, this parameter set can include multiple linear mapping parameters, i.e., different linear equations are used for conversion based on different ranges of target time variance or different combinations of material characteristics. Furthermore, the preset inference parameter set can also include nonlinear mapping functions for different material types, moisture contents, and particle morphology combinations, such as polynomial functions, exponential functions, or neural network models. These nonlinear functions can more accurately capture the complex nonlinear relationship between the target time variance and bulk density. Before practical application, these parameters and functions typically need to be calibrated through experiments, measuring the corresponding target time variance and actual bulk density under different materials, moisture contents, and particle morphologies, thereby training and optimizing them. Then, based on the inference parameters, the target time variance is converted into material bulk density. This transformation process can be a direct mathematical operation or a complex process based on lookup tables or model reasoning.

[0121] To illustrate this technical solution more clearly, a specific example is used below. Assume the dry material silo may store three materials: pulverized coal, cement, and limestone. Before the system is put into use, numerous experiments will be conducted to calibrate these three materials under different moisture contents (e.g., 0%, 5%, 10%) and different particle sizes (e.g., coarse, medium, fine). Through these experiments, a preset inference parameter set containing multiple linear mapping parameters and nonlinear mapping functions can be established. Specifically, when the system detects that the material in the dry material silo is pulverized coal, and the moisture content obtained by the humidity sensor is 5%, and it is determined to be of medium particle size through visual recognition or preset information, the system will select a conversion parameter specifically for the "pulverized coal - 5% moisture content - medium particle size" combination from the preset inference parameter set based on this information. For example, this parameter might be a specific linear equation: If the material becomes cement with a moisture content of 2% and fine particles, the system will select another set of nonlinear mapping functions for the combination of "cement - 2% moisture content - fine particles", such as a polynomial function:

[0122] .

[0123] In this way, regardless of the type, moisture content, or particle shape of the material in the dry silo, the system can select the most suitable conversion model, thereby ensuring that the calculated material bulk density has high accuracy and high reliability.

[0124] Through the above technical solution, this embodiment comprehensively considers various characteristics of the material and uses a preset inference parameter set for adaptive transformation, enabling the calculated bulk density to more accurately reflect the true state of the material. This not only improves the reliability of the detection data but also allows the method of this embodiment to be widely applied to various bulk density detection scenarios for dry silos, effectively avoiding measurement deviations caused by changes in material characteristics. This provides more reliable data support for the refined management of dry silos and the optimization of the production process.

[0125] In some embodiments, in step S106, calculating the material bulk density based on the target time variance may include, but is not limited to, the following steps:

[0126] Multiple sampling points are selected from the target energy envelope;

[0127] Calculate the time standard deviation based on the target time variance and the energy envelope value corresponding to each sampling point;

[0128] Calculate the skewness and kurtosis values ​​based on the time standard deviation;

[0129] The bulk density of the material is calculated based on the time standard deviation, skewness, kurtosis and preset conversion rules.

[0130] In some embodiments, relying solely on the target time variance to characterize the energy distribution of ultrasonic echo signals can have limitations. For example, when the energy envelopes of different materials have similar time variances but significant differences in their shape, symmetry, or tailing characteristics, a single time variance parameter may not be sufficient to distinguish these differences, thus affecting the accuracy of the bulk density calculation. Therefore, multiple sampling points can be selected from the target energy envelope. For instance, a series of discrete energy envelope values ​​can be obtained along the time axis of the target energy envelope at a certain sampling frequency or interval. These sampling points should cover the main energy distribution area of ​​the target energy envelope to ensure the representativeness of subsequent calculations. For example, equally spaced sampling, adaptive sampling, or energy threshold-based sampling methods can be used.

[0131] Then, based on the target time variance and the energy envelope value corresponding to each sampling point, the time standard deviation is calculated, allowing for further statistical quantification of the time distribution of the target energy envelope. The time standard deviation is the square root of the time variance, providing an intuitive measure of the dispersion of the energy envelope on the time axis. During the calculation, the target time variance can be used as a reference, combined with the energy envelope values ​​of each sampling point and their corresponding time positions, to accurately calculate the time standard deviation through weighted averaging or other statistical methods.

[0132] Next, based on the time standard deviation, skewness and kurtosis are calculated. Skewness measures the symmetry of the energy envelope distribution, i.e., whether it is skewed to the left or right. Positive skewness indicates a longer right tail, while negative skewness indicates a longer left tail. Kurtosis measures the sharpness or flatness of the energy envelope distribution, i.e., how concentrated its peaks are relative to its tails. High kurtosis indicates a sharper peak and thicker tails, while low kurtosis indicates a flatter distribution. These higher-order statistical moments can more precisely describe the asymmetry and concentration characteristics of the energy envelope.

[0133] Finally, the bulk density of the material is calculated based on the time standard deviation, skewness, kurtosis, and a pre-defined transformation rule. These multi-dimensional statistical characteristics can be combined and the bulk density of the material can be derived through a pre-established mapping relationship or model. The pre-defined transformation rule can be a multivariate linear regression model, a non-linear regression model, a lookup table, a neural network model, or other machine learning algorithms. This transformation rule is typically established through experimental calibration and data training on materials with known bulk densities, aiming to capture the complex relationship between the time standard deviation, skewness, kurtosis, and the bulk density of the material.

[0134] This embodiment introduces higher-order statistical moments such as time standard deviation, skewness, and kurtosis to provide a more comprehensive and detailed characterization of the shape features of the target energy envelope. Time standard deviation, as a measure of the dispersion of energy distribution, together with the target time variance, provides information on the degree of energy diffusion along the time axis. Skewness reveals the symmetry of energy distribution; for example, when material particles are unevenly distributed or agglomerated, the energy envelope of the ultrasonic echo signal may exhibit an asymmetrical tail, a feature effectively captured by skewness. Kurtosis reflects the concentration of energy distribution; for example, when material particles are uniform and densely packed, the energy envelope may be more concentrated, with a higher kurtosis; conversely, when the material is loose or contains voids, the energy distribution may be flatter, with a lower kurtosis. These additional statistical parameters quantify the shape features of the energy envelope from different dimensions, allowing the inference of material bulk density to no longer rely solely on a single dispersion index, but rather to comprehensively consider the complete form of the energy distribution.

[0135] To illustrate this technical solution more clearly, a specific example is used below. Suppose there are two different types of materials, A and B, in a dry material silo. They exhibit similar target time variances on the target energy envelope of the ultrasonic echo signal. If the bulk density is calculated solely based on the target time variance, similar bulk density values ​​may be obtained, making it impossible to accurately distinguish between the two materials. However, this embodiment allows for further calculation of the time standard deviation, skewness, and kurtosis of the target energy envelopes for both materials. For example, material A may exhibit an approximately symmetrical energy envelope distribution due to its relatively uniform particle size, with a skewness close to zero and a high kurtosis, indicating concentrated energy. Material B, on the other hand, may exhibit a significant right-skewed tail in its energy envelope due to larger particle size differences or some agglomeration, with a positive skewness and a low kurtosis, indicating a flatter energy distribution. By using these time standard deviations, skewness values, and kurtosis values ​​as input, and combining them with preset transformation rules (e.g., a trained machine learning model), the system can accurately identify the true difference in bulk density between material A and material B, even if they have similar target time variances. Therefore, this embodiment can provide more refined and accurate material bulk density detection results, effectively improving the level of intelligence in dry material silo material management.

[0136] Through the above technical solution, this embodiment, by introducing time standard deviation, skewness, and kurtosis, can more comprehensively and precisely capture the shape characteristics of the energy envelope of ultrasonic echo signals, such as their symmetry, sharpness, and tailing characteristics. These additional statistical parameters can effectively distinguish materials with similar time variances but different energy distribution patterns, thereby significantly improving the accuracy and robustness of bulk density calculation, and are particularly suitable for scenarios requiring precise detection of the bulk density of complex, non-homogeneous, or multi-component materials.

[0137] The beneficial effects of implementing the embodiments of the present invention include: the embodiments of this application first acquire the ultrasonic echo signal reflected by the material, then perform full-wave rectification and low-pass filtering on the ultrasonic echo signal to obtain the initial energy envelope, then perform noise reduction processing on the initial energy envelope to obtain the target energy envelope, and calculate the target time variance of the target energy envelope. Finally, the material bulk density is calculated based on the target time variance, thereby enabling the calculation of material bulk density by combining the energy envelope and the time variance, so as to realize bulk density detection and improve accuracy and reliability.

[0138] like Figure 2 As shown in the figure, this embodiment of the invention also provides a dry material silo bulk density detection system, including:

[0139] The signal acquisition module 201 is used to acquire the ultrasonic echo signal reflected by the material after an ultrasonic pulse is emitted into the material in the dry material silo.

[0140] The signal rectification module 202 is used to perform full-wave rectification processing on the ultrasonic echo signal, so that the negative value of the instantaneous amplitude in the ultrasonic echo signal is converted into a positive value.

[0141] The envelope extraction module 203 is used to perform low-pass filtering on the ultrasonic echo signal after full-wave rectification to obtain the initial energy envelope.

[0142] The denoising module 204 is used to denoise the initial energy envelope to obtain the target energy envelope;

[0143] The time variance calculation module 205 is used to calculate the target time variance of the target energy envelope;

[0144] The bulk density calculation module 206 is used to calculate the bulk density of materials based on the target time variance.

[0145] The content of the above method embodiments is applicable to this system embodiment. The specific functions implemented in this system embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0146] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

Claims

1. A method of detecting the bulk density of a dry material bin, characterized by, The method comprises the following steps: After emitting an ultrasonic pulse to the material in the dry material bin, an ultrasonic echo signal reflected by the material is collected; Full-wave rectification processing is performed on the ultrasonic echo signal, so that the negative value of the instantaneous amplitude in the ultrasonic echo signal is converted into a positive value; Low-pass filtering is performed on the ultrasonic echo signal after full-wave rectification processing to obtain an initial energy envelope; De-noising processing is performed on the initial energy envelope to obtain a target energy envelope; A target time variance of the target energy envelope is calculated; According to the target time variance, the bulk density of the material is calculated. The de-noising processing on the initial energy envelope to obtain a target energy envelope comprises: Segmentation processing is performed on the initial energy envelope to obtain a plurality of candidate energy regions; According to the overall average energy and the local energy fluctuation degree of the initial energy envelope, an energy threshold is determined; An energy region is selected from the plurality of candidate energy regions as a to-be-identified energy region; If the energy envelope value in the to-be-identified energy region is greater than the energy threshold, it is judged whether the duration of the to-be-identified energy region is greater than a first preset duration threshold; If the duration of the to-be-identified energy region is greater than the first preset duration threshold, it is judged whether the energy concentration degree of the to-be-identified energy region is less than a preset concentration threshold; If the energy concentration degree of the to-be-identified energy region is less than the preset concentration threshold, the to-be-identified energy region is taken as a wide secondary energy region; Energy weighted attenuation processing is performed on the wide secondary energy region in the initial energy envelope to obtain the target energy envelope. After the de-noising processing on the initial energy envelope to obtain a target energy envelope, the method further comprises: A secondary energy peak in the target energy envelope is identified, the duration of the secondary energy peak is less than a preset time window, and the amplitude of the secondary energy peak is greater than a preset amplitude threshold; If the secondary energy peak is located in the front region of the target energy envelope, the secondary energy peak is suppressed, and the target energy envelope is updated; The calculation of the target time variance of the target energy envelope comprises: A tail region after a main energy peak in the target energy envelope is identified, the amplitude of the tail region is less than the amplitude of the main energy peak, and the duration of the tail region is greater than a second preset duration threshold; According to the amplitude of the tail region and the duration of the tail region, an amplitude-time product is calculated; According to the amplitude-time product, a correction weight of the tail region is calculated; According to the energy of the tail region and the correction weight, the target time variance is calculated.

2. The method of claim 1, wherein, The calculation of the target time variance of the target energy envelope comprises: An energy concentration region in the target energy envelope is identified; A local time variance of the energy concentration region is calculated; According to the local time variance and an energy weight, a correction time variance is calculated; According to the correction time variance, the target time variance is calculated.

3. The method of claim 1, wherein, The calculation of the target time variance of the target energy envelope comprises: selecting a plurality of sampling points from the target energy envelope; calculating a centroid time of the target energy envelope according to an energy envelope value corresponding to each sampling point and a total number of sampling points; calculating a weighted sum of squares as the target time variance according to the centroid time and a point weight corresponding to each sampling point.

4. The method of claim 1, wherein, The calculation of the target time variance of the target energy envelope comprises: obtaining an amplitude and an arrival time of a main energy peak in the target energy envelope; if the amplitude of the main energy peak is less than a preset reference amplitude, and the arrival time is not the same as a preset reference time, adjusting an ultrasonic wave emission direction or a receiving gain, and updating the target energy envelope; calculating the target time variance according to the updated target energy envelope.

5. The method of claim 1, wherein, The calculation of the material bulk density according to the target time variance comprises: obtaining category information, moisture content information and particle shape information of the material; selecting corresponding inference parameters from a preset inference parameter set according to the category information, the moisture content information and the particle shape information, the preset inference parameter set comprising a plurality of linear mapping parameters or a nonlinear mapping function for different combinations of material category, moisture content and particle shape; converting the target time variance into the material bulk density according to the inference parameters.

6. The method of claim 1, wherein, The calculation of the material bulk density according to the target time variance comprises: selecting a plurality of sampling points from the target energy envelope; calculating a time standard deviation according to the target time variance and an energy envelope value corresponding to each sampling point; calculating a skewness value and a kurtosis value according to the time standard deviation; calculating the material bulk density according to the time standard deviation, the skewness value, the kurtosis value and a preset conversion rule.

7. A dry bulk storage bin bulk density detection system characterized by, It comprises: a signal acquisition module, configured to acquire an ultrasonic wave echo signal reflected by the material after emitting an ultrasonic wave pulse to the material in the dry material bin; a signal rectification module, configured to perform full-wave rectification processing on the ultrasonic wave echo signal, so that a negative value of an instantaneous amplitude in the ultrasonic wave echo signal is converted into a positive value; an envelope line extraction module, configured to perform low-pass filtering on the ultrasonic wave echo signal after full-wave rectification processing to obtain an initial energy envelope; a denoising module, configured to perform denoising processing on the initial energy envelope to obtain a target energy envelope; a time variance calculation module, configured to calculate a target time variance of the target energy envelope; a bulk density calculation module, configured to calculate a material bulk density according to the target time variance; wherein the denoising processing on the initial energy envelope to obtain the target energy envelope comprises: segmenting the initial energy envelope to obtain a plurality of candidate energy regions; determining an energy threshold according to an overall average energy and a local energy fluctuation degree of the initial energy envelope; selecting an energy region from the plurality of candidate energy regions as a to-be-identified energy region; if the energy envelope value in the to-be-identified energy region is greater than the energy threshold, determining whether a duration of the to-be-identified energy region is greater than a first preset duration threshold; If the duration of the to-be-identified energy region is greater than a first preset duration threshold, it is determined whether the energy concentration degree of the to-be-identified energy region is less than a preset concentration degree threshold; If the energy concentration degree of the to-be-identified energy region is less than the preset concentration degree threshold, the to-be-identified energy region is taken as a wide secondary energy region; performing energy weighting attenuation processing on the wide secondary energy region in the initial energy envelope to obtain the target energy envelope; After the initial energy envelope is denoised to obtain the target energy envelope, the method further includes: identifying a secondary energy peak in the target energy envelope, the duration of the secondary energy peak being less than a preset time window, and the amplitude of the secondary energy peak being greater than a preset amplitude threshold; If the secondary energy peak is located in the front region of the target energy envelope, the secondary energy peak is suppressed to update the target energy envelope; The calculation of the target time variance of the target energy envelope includes: identifying a tail region after a main energy peak in the target energy envelope, the amplitude of the tail region being less than the amplitude of the main energy peak, and the duration of the tail region being greater than a second preset duration threshold; calculating an amplitude-time product according to the amplitude of the tail region and the duration of the tail region; calculating a correction weight of the tail region according to the amplitude-time product; calculating the target time variance according to the energy of the tail region and the correction weight.

Images