On-line detection device and method for mixing uniformity of fly ash slurry

Abstract

The invention relates to the technical field of industrial slurry performance detection, in particular to a fly ash slurry mixing uniformity online detection device and method, which comprises a multi-parameter integrated sensor module, a telescopic lifting mechanical module, a pipeline bypass detection module, a signal processing and calculating unit and a human-computer interaction interface, the multi-parameter integrated sensor module integrates a microwave concentration sensor, an optical sensor and an ultrasonic sensor, and synchronously acquires various physical parameters; the telescopic lifting mechanical module controls the sensor to scan in the vertical direction of the tank body to obtain depth distribution information; the pipeline bypass detection module realizes continuous monitoring in a flowing state; the signal processing and calculating unit carries out nonlinear fusion on the multi-parameter signals based on a phase-space reconstruction technology, calculates a uniformity index, and analyzes and predicts a settlement trend through attractor dynamic characteristics; the real-time online detection of the mixing uniformity of the fly ash slurry is realized, and the problems of hysteresis, one-sidedness and subjectivity of a traditional method are solved.

Description

Technical Field

[0001] This invention relates to the field of industrial slurry performance testing technology, and in particular to an online testing device and method for the mixing uniformity of fly ash slurry. This device is suitable for evaluating the concentration consistency, particle dispersion, and stability of slurry, and can be linked with the mixing and preparation system to form a closed-loop control. Background Technology

[0002] In the preparation and use of fly ash slurry, mixing uniformity is a key indicator of its quality. Poor uniformity of the slurry can easily lead to problems such as settling blockage, pumping difficulties, and incomplete reaction. Currently, the testing of slurry uniformity is mostly done offline and manually, which has significant drawbacks:

[0003] Lag: Manual sampling and subsequent laboratory analysis (such as concentration measurement by drying and weighing, laser particle size analysis, etc.) result in severely delayed results, making it unsuitable for real-time control of the production process.

[0004] One-sidedness: Single-point sampling cannot represent the overall condition of the entire tank or pipeline, and it is especially difficult to detect stratification and settlement.

[0005] Subjectivity: Relying on experience-based judgment (such as visual inspection) lacks objective data support.

[0006] Unable to be automated: Offline detection interrupts the continuous production process, making it difficult to achieve intelligent closed-loop control.

[0007] Existing online detection methods mostly use a single parameter (such as conductivity or density) to evaluate the slurry state, which is insufficient to comprehensively reflect the mixing uniformity of the slurry. At the same time, traditional detection technologies are mostly based on simple statistical methods, lacking in-depth analysis of the dynamic characteristics of the slurry, and thus failing to provide early warning of settling trends. Summary of the Invention

[0008] The purpose of this invention is to overcome the shortcomings of existing offline detection technologies and provide a device and method that can detect the mixing uniformity of fly ash slurry in real time, online, and non-destructively, and can provide early warning of settling trends.

[0009] This invention proposes an online detection device for the mixing uniformity of fly ash slurry, comprising:

[0010] A multi-parameter integrated sensor module is used to collect physical parameter information of fly ash slurry;

[0011] A telescopic lifting mechanical module is connected to the multi-parameter integrated sensor module and is used to adjust the position of the multi-parameter integrated sensor module in the fly ash slurry;

[0012] The pipeline bypass detection module is used to monitor the flow of fly ash slurry;

[0013] The signal processing and computing unit is connected to the multi-parameter integrated sensor module and the pipeline bypass detection module, respectively. It is used to receive the signals collected by the multi-parameter integrated sensor module and the pipeline bypass detection module, perform nonlinear fusion processing on the signals based on phase space reconstruction technology, generate a uniformity index, and predict the settlement trend based on attractor dynamic feature analysis.

[0014] The system also includes a human-computer interaction interface connected to the signal processing and computing unit, used to display the uniformity index and settlement trend prediction results.

[0015] Preferably, the multi-parameter integrated sensor module includes:

[0016] Microwave concentration sensor used to measure the volume concentration of fly ash slurry;

[0017] Optical sensors are used to measure the transmittance or scattered light intensity of fly ash slurry;

[0018] And ultrasonic sensors, used to measure the sound velocity or attenuation value of fly ash slurry;

[0019] The microwave concentration sensor, the optical sensor, and the ultrasonic sensor are integrated into the same probe to ensure consistency in measurement location and time.

[0020] Preferably, the telescopic lifting mechanical module includes:

[0021] An electric actuator or servo motor is used to drive the multi-parameter integrated sensor module to move vertically;

[0022] And a position controller, used to control the multi-parameter integrated sensor module to stay at a preset depth position for a preset time to achieve vertical scanning measurement.

[0023] Preferably, the pipeline bypass detection module includes:

[0024] Bypass circulation pipeline is connected in parallel to the main process pipeline;

[0025] A window-type flow tank is installed on the bypass circulation pipeline;

[0026] And sensor probes are installed on both sides of the window-type flow pool for continuous monitoring of the flowing fly ash slurry.

[0027] Preferably, the signal processing and computing unit includes:

[0028] The data preprocessing module is used to filter, standardize, and detect anomalies in the signals collected by the multi-parameter integrated sensor module and the pipeline bypass detection module.

[0029] The phase space construction module is used to construct a high-dimensional phase space representation based on the preprocessed signal;

[0030] The feature extraction module is used to extract feature vectors characterizing the mixing state of fly ash slurry from the phase space representation;

[0031] A uniformity calculation module is used to calculate a uniformity index based on the feature vector.

[0032] It also includes an attractor analysis module for predicting settlement trends based on the dynamic characteristics of phase space trajectories.

[0033] Preferably, the phase space construction module constructs the high-dimensional phase space in the following manner:

[0034] An independent phase space is constructed for each sensor signal using the delay coordinate method, and the embedding dimension and time delay parameters are adaptively determined by the signal characteristics.

[0035] Furthermore, the correlation between the signals of each sensor is calculated based on the normalized mutual information entropy, and a unified phase space after fusion is constructed.

[0036] Preferably, the features extracted by the feature extraction module include:

[0037] The geometric characteristics of the phase space trajectory include point cloud diameter, average distance, and trajectory curvature;

[0038] The dynamic characteristics of phase space trajectories, including the Lyapunov exponent and correlation dimension;

[0039] And the statistical characteristics of the phase space point distribution, including standard deviation, skewness, and kurtosis.

[0040] Preferably, the attractor analysis module includes:

[0041] The attractor feature extraction unit is used to identify attractor structures in phase space and extract their geometric features and quantitative characteristic parameters.

[0042] The evolution trend analysis unit is used to monitor changes in attractor characteristics over time and analyze the rate and direction of change.

[0043] It also includes a multi-level anomaly detection unit, which is used to comprehensively assess settlement risk through statistical detection, trajectory bifurcation detection, and attractor deformation detection.

[0044] Preferably, the human-computer interaction interface includes:

[0045] The uniformity display module is used to display the real-time uniformity index in both numerical and graphical formats;

[0046] The parameter curve display module is used to display real-time curves and historical trends of concentration, optical signal, and viscosity;

[0047] The vertical distribution display module is used to display the parameter distribution curves in the vertical direction;

[0048] It also includes an early warning display module, which is used to issue early warning prompts when the uniformity index is lower than a preset threshold or when a settlement trend is detected.

[0049] The online detection method for the mixing uniformity of fly ash slurry based on the aforementioned device is characterized by comprising the following steps:

[0050] The installation and calibration steps include installing the multi-parameter integrated sensor module on the top of the mixing tank, connecting the pipeline bypass detection module to the main pipeline, and calibrating the sensor using standard slurry.

[0051] The steps for selecting a detection mode include choosing either a profile scanning mode or a fixed-point monitoring mode based on operational requirements.

[0052] The data acquisition steps include controlling the telescopic lifting mechanical module to perform a uniform speed scan from the liquid surface to the bottom of the tank in the profile scanning mode, or fixing the probe at a specific position in the fixed-point monitoring mode.

[0053] The data processing steps include preprocessing the acquired multi-parameter signals, constructing a high-dimensional phase space, and extracting feature vectors that characterize the mixing uniformity.

[0054] The uniformity evaluation step includes calculating a uniformity index based on the feature vector and comparing it with historical data to assess the current uniformity level.

[0055] The trend prediction step includes analyzing the dynamic changes in phase space attractor characteristics to predict slurry settling trends;

[0056] The system includes display and warning steps, such as displaying uniformity index, parameter curves, and warning information on the human-machine interface, and triggering a warning signal when the uniformity index is below a threshold or when a significant settlement trend is detected.

[0057] The beneficial effects of this invention include:

[0058] 1. Real-time online detection: Completely changes the lag of offline detection, realizes synchronous quality monitoring in the production process, and transforms production decision-making from post-analysis to real-time control.

[0059] 2. Comprehensive and objective evaluation: Combining vertical profile scanning and continuous pipeline monitoring, it can not only detect longitudinal stratification but also assess the overall macroscopic uniformity, providing multi-dimensional objective data to replace experience-based judgment.

[0060] 3. Predictive maintenance and control: Based on uniformity index and trend analysis, settlement and uniformity deterioration trends can be detected in advance, enabling predictive maintenance and early intervention. The warning time can be 30 minutes in advance, far exceeding the 5 to 10 minutes of traditional methods.

[0061] 4. System Integration and Automation: Standard signal output (4-20mA, Modbus, etc.) can be easily integrated into existing DCS or PLC control systems, providing key data support for fully automated intelligent production.

[0062] 5. Energy saving and environmental benefits: Precise control of the mixing process avoids over-mixing, reducing energy consumption by 15% to 25%. At the same time, non-destructive testing eliminates reagent consumption, making it environmentally friendly and economical. Attached Figure Description

[0063] Figure 1 This is a schematic diagram of the overall system configuration of the device of the present invention. Detailed Implementation

[0064] Please refer to Figure 1 The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be noted that the following embodiments are for illustrative purposes only and are not intended to limit the scope of the invention.

[0065] like Figure 1 As shown, the online detection device for the uniformity of fly ash slurry mixing of the present invention includes a multi-parameter integrated sensor module 1, a telescopic lifting mechanical module 2, a pipeline bypass detection module 3, a signal processing and calculation unit 4, and a human-machine interface 5.

[0066] In a preferred embodiment of the present invention, the multi-parameter integrated sensor module 1 is connected to the telescopic lifting mechanical module 2, enabling data collection at different depths in the fly ash slurry; the signal processing and calculation unit 4 is connected to both the multi-parameter integrated sensor module 1 and the pipeline bypass detection module 3, receiving and processing the collected signals; the human-machine interface 5 is connected to the signal processing and calculation unit 4, displaying the detection results and providing interactive operation. Furthermore, this system can also be connected to a mixing preparation system via a standard communication interface to form a closed-loop control.

[0067] The multi-parameter integrated sensor module 1 is the core detection unit of this invention, characterized by the simultaneous integration of three sensors based on different principles: a microwave concentration sensor, an optical sensor, and an ultrasonic sensor.

[0068] A microwave concentration sensor is used to measure the volume concentration of fly ash slurry. Preferably, a microwave sensor with an operating frequency of 2 GHz is used, with a measurement range of 0% to 60% concentration and an accuracy of ±1%. This sensor is insensitive to flow patterns, minimally affected by air bubbles, and can reflect macroscopic concentration uniformity.

[0069] The optical sensor is either an optical transmittance sensor or a laser scattering sensor, which reflects the turbidity or particle dispersion of the slurry by measuring changes in light intensity. In one embodiment of the invention, a 650nm red LED is used as the light source, in conjunction with a high-sensitivity photodiode receiver, to measure the transmitted light intensity. Local agglomeration or changes in particle distribution cause drastic fluctuations in the optical signal, thereby sensitively characterizing the microscopic dispersion uniformity.

[0070] Ultrasonic sensors indirectly reflect the apparent viscosity and density of slurry by measuring the propagation speed or attenuation of ultrasonic waves in the slurry. Preferably, an ultrasonic probe with a frequency of 1 MHz is used, and the measured sound velocity range is 1200–1600 m / s, which can be used to assess the stability of the slurry.

[0071] These three sensors are integrated into a compact probe, typically 40–60 mm in outer diameter, with a 6L stainless steel housing and a sapphire protective window, making it wear-resistant and corrosion-resistant. The synchronous data acquisition design ensures consistency in measurement location and time, avoiding the biases inherent in traditional multi-point detection.

[0072] The telescopic lifting mechanism 2 is used to vertically insert the multi-parameter integrated sensor module 1 into the mixing tank or storage tank. This module consists of an electric actuator or servo motor and a position controller.

[0073] An electric push rod or servo motor drives the probe to move vertically inside the tank. In a preferred embodiment of the invention, a servo motor system with a stroke of 2 meters is used, achieving a positioning accuracy of ±1mm and capable of withstanding a probe weight of up to 25kg.

[0074] The position controller is responsible for precisely controlling the depth position of the probe inside the tank, and supports multiple preset measurement points. Preferably, the probe is programmed to perform scanning measurements in the vertical direction, for example, stopping for 5 seconds at each measurement point every 10cm, thereby obtaining uniformity data at different depths on the longitudinal section of the entire tank, completely solving the problem of layer detection.

[0075] In addition, the position controller is equipped with limit protection and overload protection functions to ensure safe shutdown in abnormal situations (such as probe obstruction). The controller supports manual operation mode and automatic scanning mode, which can be flexibly switched according to process requirements.

[0076] The pipeline bypass detection module 3 is designed for slurry transported in pipelines, including a bypass circulation pipeline, a window-type flow tank, and a sensor probe.

[0077] The bypass circulation line is connected in parallel to the main process line, preferably using the same material as the main line, with an inner diameter typically 1 / 4 to 1 / 3 that of the main line. This design ensures that the flow conditions in the bypass are similar to those in the main line, without interfering with the main process.

[0078] The window-type flow cell is installed on the bypass circulation pipeline and employs a special design to stabilize the velocity and flow pattern of the fluid as it passes through. In one embodiment of the invention, the flow cell has an inner diameter of 80 mm and a length of 200 mm, with observation windows on both sides. The windows are made of wear-resistant sapphire material to ensure that they do not wear out during long-term use.

[0079] The sensor probe is similar to the multi-parameter integrated sensor module 1, but optimized for online flow detection. It is sealed and installed on both sides of the flow tank to achieve continuous flow detection of the slurry. This module is mainly used to evaluate the instantaneous uniformity and stability of the outlet slurry and can complement the telescopic lifting module to provide more comprehensive information.

[0080] The signal processing and calculation unit 4 is the core intelligent analysis system of this invention, including a data preprocessing module, a phase space construction module, a feature extraction module, a uniformity calculation module, and an attractor analysis module.

[0081] The data preprocessing module receives the raw signals from the sensors and performs filtering, standardization, and anomaly detection. In a preferred embodiment of the invention, the sampling frequency is set to 100Hz to ensure the capture of the rapid fluctuation characteristics of fly ash slurry.

[0082] This module first applies a third-order Butterworth low-pass filter to the raw signal, with the cutoff frequency adaptively adjusted according to the slurry flow characteristics, typically set in the range of 5–10 Hz. Then, standardization is performed, using the Z-score method to eliminate dimensional differences between signals from different sensors.

[0083] ,

[0084] in: The standardized signal value, i.e., the first... Normalized values ​​at each sampling point; The original signal value, i.e., the first... Sensor readings at each sampling point; The signal mean is calculated from data within the last 30 seconds. The standard deviation of the signal is calculated from the data within the last 30 seconds.

[0085] In practical applications, for microwave concentration sensors, The unit is volume percentage (%); for optical sensors, The unit is transmittance (%) or scattered light intensity (mV); for ultrasonic sensors, The units are sound velocity (m / s) or attenuation coefficient (dB / cm). After standardization, all signals are converted into dimensionless values, facilitating subsequent multi-parameter fusion processing.

[0086] Anomaly detection is based on statistical characteristics within a sliding window. When a signal value deviates from more than three standard deviations, it is marked as an anomaly and smoothed. For example, if a microwave concentration sensor reading suddenly jumps from 40% to 20%, exceeding the normal fluctuation range (typically ±3%), the system will mark this point as an anomaly and replace it with the average of nearby points. The processed data is timestamped and stored in a buffer for use by subsequent modules.

[0087] The phase space construction module is the first innovation of this invention. Based on the phase space reconstruction technology in chaos theory, it maps the heterogeneous data of the three sensors into a unified high-dimensional phase space.

[0088] First, an independent phase space is constructed for each sensor signal. Using the delay coordinate method, taking a microwave concentration sensor as an example, its phase space points are represented as follows:

[0089] ,

[0090] in: For a point in phase space, it is a dimensional vector; For the first time series One data point; This is a time delay, expressed in units of the number of sampling points. The embedding dimension represents the dimension of the phase space. In the specific application of fly ash slurry monitoring, the time delay is determined by the first zero point of the signal autocorrelation function. For medium-concentration fly ash slurry with a concentration of 30%–50%, the typical microwave signal... The value is approximately 20 sampling points (0.2 seconds), a choice based on the rheological properties and fluctuation frequency of the fly ash slurry. The embedding dimension *m* is determined using the pseudo-nearest neighbor method, set to 6 for microwave signals, 7 for optical signals, and 5 for ultrasonic signals. The choice of dimension takes into account the different complexities of the physical processes involved in capturing each signal; for example, optical signals are more complex due to the influence of particle micro-distribution and require a higher dimension for full characterization.

[0091] Subsequently, the topological mapping relationship between the sensors is calculated, and a 3×3 correlation matrix is ​​constructed. :

[0092] ,

[0093] in: For sensor signals and The normalized mutual information value between them, ranging from 0 to 1; Mutual information represents the statistical correlation between two signals; and These are the entropy values ​​of each signal, representing the uncertainty of the signal.

[0094] In the detection of fly ash slurry homogeneity, the correlation matrix reflects the inherent relationship between the physical characteristics captured by different sensors. For example, in a homogeneously mixed slurry, the microwave concentration signal and the optical transmittance signal typically exhibit a strong negative correlation (…). The correlation is approximately 0.7-0.8, but it decreases significantly under non-uniform conditions. (Reduced to 0.3-0.5), becoming an important indicator for detecting non-uniformity.

[0095] Finally, a fusion phase space is constructed based on the correlation matrix, and a weighted combination method is used:

[0096] ,

[0097] in: For points in the fused phase space, it is a high-dimensional vector; Let j be the phase space point of the j-th sensor; The weighting coefficients are dynamically adjusted based on signal quality and correlation. In practical applications, for fly ash slurry with a concentration of 30%–50%, the microwave signal weighting... It is usually set to 0.4-0.5 because it reliably reflects macroscopic concentration; optical signal weighting The value is 0.3-0.4, due to its sensitivity to microscopic dispersion; ultrasonic signal weighting The weight is set at 0.2-0.3 because it provides supplementary information on slurry stability. When an anomaly is detected in a sensor (such as optical window contamination), the system automatically reduces its weight to ensure the reliability of the fusion results.

[0098] The feature extraction module extracts feature vectors characterizing the mixing state of fly ash slurry from the phase space representation, including geometric features, kinetic properties, and statistical features.

[0099] Geometric features include the diameter of the phase space point cloud. Average distance and trajectory curvature :

[0100] ,

[0101] ,

[0102] ,

[0103] in: The maximum diameter of the point cloud in phase space represents the limiting range of system state changes; The average Euclidean distance between point pairs reflects the degree of dispersion of the system state; Represents the first phase space Each point is a multidimensional vector; The total number of points is usually taken from the data points within the last 30 seconds, approximately 3000 points; and These are the first and second derivatives of the trajectory at that point, representing the velocity and acceleration of the system state change; The average curvature characterizes the complexity of the trajectory; Represents Euclidean distance. This represents the cross product of vectors.

[0104] In fly ash slurry testing, well-mixed slurries typically exhibit a smaller point cloud diameter. (For a 50% concentration slurry, the typical value is approximately 0.2-0.3, after dimensionless conversion) and has a lower curvature. (Typical values ​​are approximately 0.1-0.2); while heterogeneous slurries exhibit larger values. Values ​​(0.5-0.8) and higher The value (0.3-0.5) reflects the instability of the system state.

[0105] Dynamic characteristics include the maximum Lyapunov exponent and correlation dimension The maximum Lyapunov exponent is calculated as follows:

[0106] ,

[0107] in: The maximum Lyapunov exponent characterizes the degree of chaos and the difficulty of prediction of the system. This represents the distance between two neighboring points in the phase space in the initial state. For the time elapsed The distance between these two points; Time, in seconds; It is the natural logarithm.

[0108] In practical implementation, the evolutionary trajectories of neighboring point pairs in phase space are calculated to estimate... Value. For fly ash slurry, a uniform and stable slurry typically exhibits a smaller [value]. The values ​​are approximately 0.05-0.1 for the uniform slurry and larger values ​​for the heterogeneous slurry (0.2-0.4), indicating that the system is more unpredictable.

[0109] Correlation dimension Calculated via correlation integral:

[0110] ,

[0111] ,

[0112] in: The correlation integral represents the distance in phase space less than 1 / 2. Point-to-point ratio; This is the Heaviside step function, which has a value of 1 when the parameter is positive and a value of 0 when the parameter is negative. The set distance threshold; The correlation dimension represents the complexity of the system.

[0113] In fly ash slurry monitoring The values ​​are typically chosen from 1% to 50% of the point cloud diameter, using 10 to 20 equally spaced values. right The slope as Estimated value. Homogeneous slurry. Typically lower (1.5-2.5), reflecting a more regular system; non-uniform slurry... A higher value (2.5-3.5) indicates a more complex system.

[0114] Statistical characteristics include the standard deviation of the phase space point distribution. skewness and kurtosis These parameters reflect the statistical characteristics of the slurry mixing state. These parameters are calculated for each phase space dimension, and then the average value is taken as the overall characteristic.

[0115] All extracted features are combined into an 18-dimensional feature vector, comprehensively describing the current mixing state of the fly ash slurry. The feature extraction frequency is 1Hz, ensuring timely system response without excessive consumption of computational resources.

[0116] The uniformity calculation module calculates the uniformity index based on the extracted feature vectors, using a nonlinear mapping method:

[0117] ,

[0118] in: It is the uniformity index, ranging from 0 to 1 (or 0% to 100%). It is the i-th component of the eigenvector; This is the corresponding nonlinear transformation function; These are the weighting coefficients; The number of features is 18 in this embodiment.

[0119] In practical applications, nonlinear transformation functions The Sigmoid function is typically used to normalize different features.

[0120] ,

[0121] in: To adjust the slope parameter and control the sensitivity of the response; The threshold parameter represents the expected value or critical point of the feature; is the base of the natural logarithm.

[0122] Typical values ​​for various parameters used in testing the uniformity of fly ash slurry are as follows: For the phase space point cloud diameter... , Approximately 10, Approximately 0.3; for the maximum Lyapunov exponent , Approximately 15, Approximately 0.15; for correlation dimension , Approximately 5, Approximately 2.5. These parameters were determined based on the analysis of a large amount of experimental data and can be fine-tuned for different concentrations and types of fly ash slurry.

[0123] The weighting coefficients α were determined through extensive experimental data analysis, reflecting the contribution of each feature to uniformity. Preferably, the weights for geometric features range from 0.3 to 0.4, for kinetic characteristics from 0.4 to 0.5, and for statistical features from 0.2 to 0.3. For example, for a 50% concentration fly ash slurry, the weight α for the point cloud diameter D is approximately 0.15, the weight α for the maximum Lyapunov exponent λmax is approximately 0.2, and the weight αD for the correlation dimension D² is approximately 0.12.

[0124] The uniformity index is updated every 3 seconds, and a 95% confidence interval is calculated to assess the reliability of the measurement. In industrial applications, multiple warning thresholds are set: when the index is below 0.9, the system generates a prompt message suggesting operators monitor the mixing status; below 0.85, a mild warning is triggered, automatically increasing the stirring speed by 10%–15%; below 0.8, a moderate warning is issued, increasing the stirring speed by 15%–20% and activating the auxiliary stirring device; below 0.75, a severe warning is triggered, increasing the stirring speed to maximum, adjusting the circulation pump flow, and stopping downstream delivery if necessary. These thresholds are set based on extensive industrial experience and effectively prevent problems caused by uneven slurry mixing.

[0125] The attractor analysis module is the second innovation of this invention. Based on the attractor analysis method in chaos theory, it achieves early prediction of sedimentation trends by monitoring the dynamic changes of attractors formed in the phase space of the fly ash slurry system. This module includes an attractor feature extraction unit 1, an evolution trend analysis unit 2, and a multi-level anomaly detection unit 3.

[0126] Attractor feature extraction unit 1 uses phase space data within a 5-minute time window to identify the attractor structure and extract its geometric morphological features and quantitative characteristic parameters. Geometric morphological features include attractor volume V, surface area A, principal axis direction u, and eccentricity e.

[0127] ,

[0128] ,

[0129] in: Let be the attractor volume, representing the extent of the system's state space covered; For attractor surface area; The phase space region occupied by the attractor; The density function of a point in phase space; For the regional boundary; It is a micro-element of area; Indicates in the region Volume integral over; This represents the area integral over the boundary of the region.

[0130] In practical calculations, a gridded method is used to estimate volume and surface area. For a typical fly ash slurry system, the attractor volume in a homogeneous state is small (approximately 0.1-0.2 after normalization) and its shape is relatively regular (eccentricity). Approaching 1); however, in the non-uniform state, the attractor volume increases (0.3-0.5) and its shape is stretched and deformed ( (Increased to 1.5-2.5). This change reflects the transition of the system from a steady state to an unstable state and is an early signal of a settling trend.

[0131] Quantitative characteristic parameters include fractal dimension. Information entropy and phase space point density distribution The fractal dimension is calculated using box counting:

[0132] ,

[0133] in: is the fractal dimension, which characterizes the complexity of the attractor structure; The required side length to cover the attractor is The number of hypercubes; For the size of the box; It is the natural logarithm.

[0134] In fly ash slurry monitoring applications Typically, 10 to 15 values ​​are selected from the range of 1 / 100 to 1 / 10 of the point cloud diameter for calculation. right The slope as Estimated value. Homogeneous slurry. The value is typically in the range of 1.2-1.8, but when the slurry is uneven or shows a settling tendency, It will increase to 1.8-2.5, indicating that the system's state space structure is more complex.

[0135] Evolution trend analysis unit 2 uses a sliding time window to monitor changes in attractor characteristics over time and calculates the morphological deviation rate between consecutive windows. :

[0136] ,

[0137] in: The morphological deviation rate characterizes the rate of change in the attractor's morphology. The first attractor Features (such as volume) Surface area fractal dimension wait); The total number of features is 6 in this embodiment; The current time; This is the width of the time window, typically set to 15–30 minutes.

[0138] In the monitoring of fly ash slurry settling, the morphological deviation rate This is a key indicator. When the slurry is in a stable and homogeneous state, It usually remains below 0.05; when a slight settling trend begins to appear, It will slowly rise to 0.05-0.1; when the settling trend is obvious, It will rise rapidly to 0.1-0.15; exceeding 0.15 usually indicates that settlement has begun to accelerate. This can be achieved through monitoring... The system can detect signs of subsidence in early stages that are undetectable by traditional methods.

[0139] The multi-level anomaly detection unit 3 comprehensively applies a three-layer detection strategy: the first layer is statistical anomaly detection, based on the coefficient of variation of concentration, transmittance, and viscosity; the second layer is phase space trajectory bifurcation detection; and the third layer is attractor deformation rate detection. The results of the three layers are fused through a decision tree structure to generate a sedimentation risk index. :

[0140] ,

[0141] in: The subsidence risk index ranges from 0 to 1. , , These are the risk indices for three layers of detection, all ranging from 0 to 1. , , For the corresponding weights, satisfying Typically set to 0.2, 0.3, or 0.5, this reflects higher reliability for high-level detection. In practical applications, for fly ash slurry, the first layer... Based on traditional statistical methods, when the coefficient of variation of concentration in the vertical direction exceeds 0.05, Start to increase; second layer Detecting bifurcation points in the phase space trajectory provides early signals characterizing system instability; third layer Direct correlation morphological deviation rate ,when When it exceeds 0.1, It grows rapidly.

[0142] Settlement risk index A threshold of 0.3 triggers an alert, 0.5 triggers a mild warning, 0.7 triggers a moderate warning, and 0.85 triggers a severe warning. In practical applications, for 50% concentration fly ash slurry, the system can predict the settling trend 30 minutes in advance, significantly earlier than the 5-10 minutes of traditional methods, providing ample time for intervention.

[0143] The human-computer interaction interface 5 includes a uniformity display module, a parameter curve display module, a vertical distribution display module, and an early warning display module.

[0144] The uniformity display module displays the real-time uniformity index in digital and graphical form (such as on a dashboard), with a value range of 0% to 100%. The green area represents 90% to 100% (normal), the yellow area represents 80% to 90% (caution), and the red area represents below 80% (warning).

[0145] The parameter curve display module displays real-time curves and historical trends of concentration, optical signal, and viscosity. The time span can be selected as 5 minutes, 30 minutes, 2 hours, and 8 hours, which is convenient for operators to analyze long-term trends.

[0146] The vertical distribution display module shows the parameter distribution curves in the vertical direction, i.e., the concentration profile, which visually displays the concentration distribution at different depths and helps identify stratification. In an ideal state of uniform mixing, the curve should be approximately vertical; while a sloping curve or the appearance of an inflection point indicates the risk of stratification.

[0147] The warning display module issues a warning when the uniformity index falls below a preset threshold or when a settlement trend is detected, including an audible and visual alarm and a text description. The warnings are divided into four levels: alert (yellow text), mild warning (yellow flashing + sound), moderate warning (orange flashing + sound), and severe warning (red flashing + continuous sound).

[0148] In addition, the human-machine interface also provides an operation control area where parameters such as detection mode, sampling frequency, and alarm threshold can be set, and touch screen operation and password protection functions are supported.

[0149] The online detection method for the mixing uniformity of fly ash slurry based on the above-mentioned device includes the following steps:

[0150] The multi-parameter integrated sensor module 1 is installed on the top of the mixing tank, with the sensor probe extending into the slurry through a flange or manhole. The pipeline bypass detection module 3 is connected to the main pipeline to ensure stable bypass circulation flow. The sensor is calibrated at multiple points using standard slurries of known concentrations (e.g., 30%, 40%, 50%) to establish the correspondence between the sensor signal and the actual concentration.

[0151] Preferably, the calibration process is controlled by an automated program, collecting at least 100 data points for each concentration point, taking the average value to establish a calibration curve, and ensuring that the measurement accuracy is within ±1%.

[0152] Select the detection mode according to operational requirements: profile scanning mode or fixed-point monitoring mode.

[0153] The profile scanning mode is suitable for the initial mixing stage of the tank and key quality control points, acquiring comprehensive stratification information through vertical scanning. The fixed-point monitoring mode is suitable for the stable operation stage, with sensors fixed at key locations (usually the middle of the tank or the outlet) to continuously monitor changes in uniformity.

[0154] Preferably, the system can be set to automatically perform profile scanning at regular intervals (e.g., once every 2 hours), while maintaining fixed-point monitoring during normal times, thus balancing comprehensiveness and real-time performance.

[0155] In profile scanning mode, the telescopic lifting mechanical module 2 drives the sensor probe to perform a uniform scanning motion from the liquid surface to the bottom of the tank. The scanning speed is typically set to 5–10 cm / s, and data is collected at preset positions (e.g., every 10 cm) for 5 seconds.

[0156] In fixed-point monitoring mode, the probe is fixed at a specific location (e.g., 1 / 3 of the height from the bottom of the tank) for continuous monitoring. The sampling frequency is 100Hz, and the uniformity index is calculated every 3 seconds.

[0157] Preferably, the system automatically records the collection location, timestamp, and environmental conditions (such as temperature) to facilitate subsequent data analysis and comparison.

[0158] The signal processing and computation unit 4 receives the acquired multi-parameter signals and first performs preprocessing, including filtering, normalization, and anomaly detection. Then, it constructs a high-dimensional phase space and extracts feature vectors representing the homogeneity of the mixture.

[0159] During phase space construction, the system dynamically adjusts the embedding dimension and time delay parameters to adapt to different slurry properties. The feature extraction process focuses on the geometric morphology, dynamic characteristics, and statistical distribution of phase space trajectories, generating an 18-dimensional feature vector.

[0160] Preferably, the system adopts a parallel computing architecture, with data preprocessing completed on the FPGA and phase space construction and feature extraction executed on a dual-core ARM processor to ensure real-time processing capabilities.

[0161] Based on the extracted feature vectors, a uniformity index is calculated and compared with historical data to assess the current level of uniformity. The uniformity index H ranges from 0 to 1 (or 0% to 100%), with higher values ​​indicating better uniformity.

[0162] The system also calculates a short-term volatility index, namely the standard deviation of the uniformity index over 30 seconds, which reflects the stability of the mixture. When the volatility index exceeds 0.05, it indicates that the mixing process is unstable, and attention should be paid even if the uniformity index is high.

[0163] Preferably, the system establishes a historical database for uniformity evaluation, recording typical uniformity levels under different working conditions, providing a reference benchmark for current evaluation.

[0164] The system analyzes the dynamic changes in the characteristics of phase space attractors to predict slurry settling trends. It identifies early signs of settling by monitoring minute changes in attractor volume, surface area, and morphological parameters.

[0165] When attractor morphology deviation rate is detected If the value continues to increase and exceeds the threshold of 0.15, the system determines that there is a risk of settlement, calculates the risk level and estimates the time of occurrence.

[0166] Preferably, the system combines multi-level anomaly detection results to comprehensively assess settlement risk, ensuring the accuracy and timeliness of early warnings. Early warning lead times are categorized into three types based on settlement speed: slow (>60 minutes), moderate (30–60 minutes), and rapid (<30 minutes).

[0167] The uniformity index, parameter curves, vertical distribution map, and early warning information are displayed in real time on the human-computer interaction interface 5. When the uniformity index is lower than the set threshold (usually four levels: 0.9, 0.85, 0.8, and 0.75) or a significant settlement trend is detected, the corresponding level of early warning signal is triggered.

[0168] The warning information includes the current uniformity, estimated settling time, risk level, and recommended measures, such as increasing the stirring speed, extending the mixing time, or adjusting the pump speed.

[0169] Preferably, the system can send early warning signals and control suggestions to the mixing control system via a standard communication protocol (such as Modbus RTU) to achieve automatic intervention and form a closed-loop control. For example, when the uniformity index is below 0.85, the stirring speed is automatically increased by 10% to 15%; when it is below 0.8, the auxiliary stirring device is started simultaneously.

[0170] The following is a specific engineering application case to illustrate the actual effect of the present invention.

[0171] The device of this invention is used in the slurry conveying system of a power plant's ash disposal area to detect the uniformity of fly ash slurry. The system configuration includes:

[0172] 1. Multi-parameter integrated sensor module: integrates a microwave concentration sensor from Hydac (Germany), a fiber optic transmittance sensor from Omega (USA), and an ultrasonic probe from Amin (Japan).

[0173] 2. Telescopic module: 2-meter travel distance, servo motor driven, positioning accuracy ±1mm.

[0174] 3. Signal processing unit: Based on ARM Cortex-M7 processor, with built-in algorithms to calculate the coefficient of variation and uniformity index.

[0175] 4. Human-machine interface: 10-inch touchscreen, displaying data and curves in real time.

[0176] In actual operation, technicians set up an automatic full-tank scan every 2 hours, and routine fixed-point monitoring is performed. The system generates uniformity reports and historical trend charts, providing data support for process optimization.

[0177] Implementation results: Uniformity increased to over 95%, sampling frequency decreased from 3-5 times per shift to once per day, the average number of blockage accidents per year decreased from 5-7 to zero, mixed power consumption decreased by 21.3%, pumping power consumption decreased by 9.6%, and the investment payback period was 8-10 months.

[0178] This invention provides an online detection device and method for the mixing uniformity of fly ash slurry. Through multi-parameter integrated sensing, vertical scanning detection, and innovative signal processing algorithms, it achieves real-time and comprehensive evaluation of slurry uniformity and early warning of settling trends. Compared with traditional technologies, this invention has advantages such as real-time detection, comprehensive evaluation, predictive foresight, and convenient integration, providing an efficient quality monitoring method for industrial slurry treatment processes.

[0179] This invention allows operators to intuitively understand the mixing state of the slurry, adjust process parameters in a timely manner, and avoid adverse operating conditions. Simultaneously, the system can form a closed-loop control with the mixing and preparation device, achieving fully automated intelligent production, improving product quality, reducing energy consumption, and extending equipment lifespan.

[0180] The embodiments described above are merely illustrative of specific implementations of the present invention, and while the descriptions are detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention.

Claims

1. A device for on-line detection of fly ash slurry mixture uniformity, characterized in that, The application relates to a multi-parameter integrated sensor module for collecting physical parameter information of fly ash slurry, a telescopic lifting mechanical module connected with the multi-parameter integrated sensor module for adjusting the position of the multi-parameter integrated sensor module in the fly ash slurry, a pipeline bypass detection module for monitoring the fly ash slurry in a flowing state, a signal processing and calculation unit connected with the multi-parameter integrated sensor module and the pipeline bypass detection module for receiving signals collected by the multi-parameter integrated sensor module and the pipeline bypass detection module, performing nonlinear fusion processing on the signals based on phase space reconstruction technology, generating a uniformity index, and performing settlement trend prediction based on attractor dynamic characteristic analysis, and a man-machine interaction interface connected with the signal processing and calculation unit for displaying the uniformity index and the settlement trend prediction result. The multi-parameter integrated sensor module comprises a microwave concentration sensor for measuring the volume concentration of the fly ash slurry, an optical sensor for measuring the transmittance or scattered light intensity of the fly ash slurry, and an ultrasonic sensor for measuring the sound velocity or attenuation value of the fly ash slurry, wherein the microwave concentration sensor, the optical sensor and the ultrasonic sensor are integrated in the same probe to ensure consistency in measurement position and time. The telescopic lifting mechanical module comprises an electric push rod or a servo motor for driving the multi-parameter integrated sensor module to vertically move, and a position controller for controlling the multi-parameter integrated sensor module to stay at a preset depth position for a preset time, thereby realizing vertical scanning type measurement. The pipeline bypass detection module comprises a bypass circulation pipeline connected in parallel with a main flow pipeline, a window type flow cell installed on the bypass circulation pipeline, and sensor probes installed on both sides of the window type flow cell for continuously monitoring the fly ash slurry flowing therethrough. The signal processing and calculation unit comprises a data preprocessing module for filtering, standardizing and performing anomaly detection on the signals collected by the multi-parameter integrated sensor module and the pipeline bypass detection module, a phase space construction module for constructing a high-dimensional phase space representation based on the preprocessed signals, a feature extraction module for extracting a feature vector representing the mixing state of the fly ash slurry from the phase space representation, a uniformity calculation module for calculating a uniformity index based on the feature vector, and an attractor analysis module for performing settlement trend prediction based on the dynamic characteristics of the phase space trajectory. The phase space construction module constructs the high-dimensional phase space by adopting a delay coordinate method to construct independent phase spaces for the sensor signals, adaptively determining embedding dimension and time delay parameters according to signal characteristics, and calculating the correlation between the sensor signals based on normalized mutual information entropy to construct a unified phase space after fusion.

2. The apparatus of claim 1, wherein, The feature extraction module extracts features including geometric characteristics of the phase space trajectory, such as point cloud diameter, average distance and trajectory curvature, dynamic characteristics of the phase space trajectory, such as Lyapunov exponent and correlation dimension, and statistical characteristics of the phase space point distribution, such as standard deviation, skewness and kurtosis. ​ ​ ​ 3. The apparatus of claim 1, wherein, ​ ​ 4. The apparatus of claim 1, wherein, ​ ​ ​ 5. The apparatus of claim 1, wherein, ​ ​ ​ ​ ​ 6. The apparatus of claim 5, wherein, ​ ​ 7. The apparatus of claim 5, wherein, ​ ​ ​ 8. The apparatus of claim 5, wherein, The attractor analysis module comprises: An attractor feature extraction unit for identifying attractor structures in the phase space and extracting their geometric shape features and quantitative characteristic parameters; An evolution trend analysis unit for monitoring changes in attractor features over time, analyzing change rates and directions; and a multi-level anomaly detection unit for comprehensive evaluation of settling risks through statistical detection, trajectory bifurcation detection, and attractor deformation detection.

9. The apparatus of claim 1, wherein, The human-computer interaction interface comprises: A uniformity display module for displaying real-time uniformity indexes in numerical and graphical ways; A parameter curve display module for displaying real-time curves and historical trends of concentration, optical signal, and viscosity; A vertical distribution display module for displaying parameter distribution curves in the vertical direction; and a warning display module for issuing warning prompts when the uniformity index is below a preset threshold or a settling trend is detected.

10. A method for on-line detection of fly ash slurry mixture uniformity, using the device according to any one of claims 1-9, characterized in that, The method comprises the following steps: An installation and calibration step comprising installing the multi-parameter integrated sensor module on the top of the mixing tank, connecting the pipeline bypass detection module to the main pipeline, and calibrating the sensor using a standard slurry; A detection mode selection step comprising selecting a profile scanning mode or a fixed-point monitoring mode according to operational requirements; A data acquisition step comprising controlling the telescopic lifting mechanical module to perform uniform scanning from the liquid surface to the tank bottom in the profile scanning mode, or fixing the probe at a specific position in the fixed-point monitoring mode; A data processing step comprising preprocessing the acquired multi-parameter signals, constructing a high-dimensional phase space, and extracting feature vectors representing mixing uniformity; A uniformity evaluation step comprising calculating a uniformity index based on the feature vectors and comparing it with historical data to evaluate the current uniformity level; A trend prediction step comprising analyzing dynamic changes in attractor features in the phase space and predicting slurry settling trends; And a display and warning step comprising displaying uniformity indexes, parameter curves, and warning information on the human-computer interaction interface and triggering a warning signal when the uniformity index is below a threshold or a significant settling trend is detected.

Images