Performance detection method, device, system, terminal equipment and storage medium for non-Newtonian fluid

Abstract

The present application relates to the technical field of non-Newtonian fluid performance detection. The present application discloses a non-Newtonian fluid performance detection method, apparatus, system, terminal device and storage medium, which can effectively improve the accuracy of detecting the performance parameters of non-Newtonian fluids. The non-Newtonian fluid performance detection method includes obtaining a reference height of the surface of the non-Newtonian fluid when the surface of the non-Newtonian fluid is in a stationary state; obtaining a height data set of the surface of the non-Newtonian fluid based on a time series when the surface of the non-Newtonian fluid is in a non-stationary state; constructing a time series curve based on the reference height and the height data set; performing feature extraction processing on the time series curve to obtain at least one feature parameter; and detecting and processing the at least one feature parameter using a preset algorithm to obtain the performance parameters of the non-Newtonian fluid.

Description

Technical Field

[0001] The present application relates to the technical field of non-Newtonian fluid performance detection technology. More specifically, the present application relates to a non-Newtonian fluid performance detection method, apparatus, system, terminal device and storage medium. Background Art

[0002] Traditional methods for testing the performance of non-Newtonian fluids measure the vibration damping of a vibrating converter; based on this vibration damping, performance parameters (such as yield stress) of the non-Newtonian fluid are directly estimated (or detected). However, this vibration damping is influenced by the structure and material properties of the vibrating converter itself, as well as fluid coupling effects, making it difficult to reflect the critical state of the non-Newtonian fluid, resulting in inaccurate measurement of non-Newtonian fluid performance parameters. Furthermore, temperature fluctuations in non-Newtonian fluids can induce shear thinning or thickening, affecting the accuracy of vibration damping measurements and further resulting in inaccurate measurement of non-Newtonian fluid performance parameters. Summary of the Invention

[0003] The purpose of the embodiments of the present application is to provide a method, apparatus, system, terminal device, and storage medium for detecting the performance parameters of non-Newtonian fluids, which can effectively improve the accuracy of detecting the performance parameters of non-Newtonian fluids. The embodiments of the present application are mainly achieved through the following technical solutions:

[0004] A first aspect of an embodiment of the present application provides a method for detecting properties of a non-Newtonian fluid, comprising:

[0005] When the surface of the non-Newtonian fluid is in a stationary state, obtaining a reference height of the surface of the non-Newtonian fluid;

[0006] When the surface of the non-Newtonian fluid is in a non-stationary state, acquiring a height dataset of the surface of the non-Newtonian fluid based on a time series;

[0007] constructing a timing curve based on the reference altitude and the altitude data set;

[0008] Performing feature extraction processing on the timing curve to obtain at least one feature parameter;

[0009] The at least one characteristic parameter is detected and processed using a preset algorithm to obtain performance parameters of the non-Newtonian fluid.

[0010] According to one embodiment of the present application, the step of constructing a timing curve based on the reference height and the height data set includes:

[0011] Calculating the difference between the reference height and each height data in the height data set to obtain target height change data corresponding to each height data;

[0012] The time series curve is constructed based on all target height change data.

[0013] According to one embodiment of the present application, the step of detecting and processing the at least one characteristic parameter using a preset algorithm to obtain the performance parameter of the non-Newtonian fluid includes:

[0014] Using the yield stress model of the preset algorithm to detect and process the at least one characteristic parameter to obtain a yield stress parameter of the performance parameter;

[0015] The plastic viscosity model of the preset algorithm is used to detect and process the at least one characteristic parameter to obtain the plastic viscosity parameter of the performance parameter.

[0016] According to one embodiment of the present application, the step of detecting and processing the at least one characteristic parameter using the yield stress model of the preset algorithm to obtain the yield stress parameter of the performance parameter includes:

[0017] performing feature extraction processing on the at least one characteristic parameter using a first hidden layer of the yield stress model to obtain a first characteristic vector;

[0018] Performing nonlinear transformation processing on the first eigenvector using a first target activation function of the yield stress model to obtain a second eigenvector;

[0019] performing feature extraction processing on the second eigenvector using the second hidden layer of the yield stress model to obtain a third eigenvector;

[0020] Performing a nonlinear transformation on the third eigenvector using the first target activation function to obtain a fourth eigenvector;

[0021] performing feature extraction processing on the fourth eigenvector using the third hidden layer of the yield stress model to obtain a fifth eigenvector;

[0022] The first target activation function is used to perform nonlinear transformation processing on the fifth eigenvector to obtain the yield stress parameter.

[0023] According to one embodiment of the present application, the method for detecting the performance of a non-Newtonian fluid further includes a step of training the yield stress model, and the step of training the yield stress model includes:

[0024] Acquire a historical data set and a first true label set, wherein each first true label in the first true label set has a one-to-one correspondence with one of the historical data in the historical data set;

[0025] Using the original yield stress model to detect and process target historical data to obtain a first prediction result, the target historical data being any historical data in the historical data set;

[0026] Constructing a first loss function based on the first prediction result and a first true label corresponding to the target historical data;

[0027] The first model parameter of the original yield stress model is adjusted based on the first loss function and the first preset constraint item to form the yield stress model.

[0028] According to one embodiment of the present application, the yield stress model training step further includes:

[0029] Using a target evaluation method to evaluate the yield stress model to obtain a first evaluation result;

[0030] A second model parameter of the yield stress model is adjusted based on the first evaluation result.

[0031] A second aspect of the embodiments of the present application provides a device for detecting properties of a non-Newtonian fluid, comprising:

[0032] A reference height acquisition module, configured to acquire a reference height of the surface of the non-Newtonian fluid when the surface of the non-Newtonian fluid is in a stationary state;

[0033] a height dataset acquisition module, configured to acquire a height dataset of the surface of the non-Newtonian fluid based on a time sequence when the surface of the non-Newtonian fluid is in a non-stationary state;

[0034] a timing curve construction module, configured to construct a timing curve based on the reference altitude and the altitude data set;

[0035] A feature extraction module, configured to perform feature extraction processing on the timing curve to obtain at least one feature parameter;

[0036] The performance parameter acquisition module is used to detect and process the at least one characteristic parameter using a preset algorithm to obtain the performance parameter of the non-Newtonian fluid.

[0037] A third aspect of an embodiment of the present application provides a non-Newtonian fluid performance detection system, comprising a laser ranging module, an external force application module, a data acquisition module, and a terminal device, wherein the terminal device comprises a processor and a memory, the processor being communicatively connected to the laser ranging module, the external force application module, and the data acquisition module, the memory being used to store a computer program, and the processor being used to call and run the computer program stored in the memory to execute the steps of the non-Newtonian fluid performance detection method provided in the first aspect of the embodiment of the present application.

[0038] In a fourth aspect of an embodiment of the present application, a terminal device is provided, comprising: a processor and a memory, the memory being used to store a computer program, the processor being used to call and run the computer program stored in the memory, and executing the steps of the non-Newtonian fluid performance detection method provided in the first aspect of the embodiment of the present application.

[0039] In a fifth aspect of an embodiment of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium is used to store a computer program, and the computer program enables a computer to execute the steps of the non-Newtonian fluid performance detection method provided in the first aspect of the embodiment of the present application.

[0040] The beneficial effects of the embodiments of the present application include:

[0041] The embodiments of the present application obtain a reference height of the surface of a non-Newtonian fluid when the surface of the non-Newtonian fluid is in a stationary state; obtain a height dataset of the surface of the non-Newtonian fluid based on a time series when the surface of the non-Newtonian fluid is in a non-stationary state; construct a time series curve based on the reference height and the height dataset; perform feature extraction processing on the time series curve to obtain at least one characteristic parameter; and detect and process the at least one characteristic parameter using a preset algorithm to obtain performance parameters of the non-Newtonian fluid. Compared with existing technologies that have difficulty reflecting the critical state of non-Newtonian fluids, the embodiments of the present application use a reference height and a height dataset of the surface of the non-Newtonian fluid, which are directly related to the critical state of the non-Newtonian fluid, to detect performance parameters, thereby improving the accuracy of detecting performance parameters of non-Newtonian fluids. Furthermore, the reference height and height dataset of the surface of the non-Newtonian fluid are not subject to the influence of the structure, material properties, and fluid coupling effects of additional equipment, thereby improving the accuracy of detecting performance parameters of non-Newtonian fluids. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the conventional technology, the following briefly introduces the drawings required for use in the embodiments or the conventional technology descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0043] Figure 1 A flow chart of a method for detecting the performance of a non-Newtonian fluid according to the present application in some embodiments;

[0044] Figure 2 This is a principle block diagram of the non-Newtonian fluid performance detection device of the present application in some embodiments;

[0045] Figure 3 This is a principle block diagram of the non-Newtonian fluid performance detection system of the present application in some embodiments;

[0046] Figure 4 This is a principle block diagram of the terminal device of the present application in some embodiments. DETAILED DESCRIPTION

[0047] To make the above-mentioned objects, features, and advantages of the present application more clearly understood, the specific embodiments of the present application are described in detail below with reference to the accompanying drawings. The following description sets forth many specific details to facilitate a full understanding of the present application. However, the present application can be implemented in many other ways than those described herein, and those skilled in the art can make similar improvements without violating the scope of the present application. Therefore, the present application is not limited to the specific embodiments disclosed below.

[0048] It should be noted that the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of such features. In the description of this application, "plurality" means at least two, for example, two, three, etc., unless otherwise specifically defined.

[0049] The terms "exemplary" or "for example" are used to indicate an example, illustration, or description. Any embodiment or design described as "exemplary" or "for example" in the embodiments of this application should not be construed as being preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0050] The terms "comprises," "comprising," or any other variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, product, or apparatus.

[0051] Unless otherwise defined, all technical and scientific terms used in the specification of this application have the same meaning as commonly understood by those skilled in the art to which this application belongs. The terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application. The term "and / or" used in the specification of this application includes any and all combinations of one or more of the relevant listed items.

[0052] The specific implementation of this application is further described below with reference to the accompanying drawings.

[0053] refer to Figure 1 The figure is a flow chart of a method for detecting the performance of a non-Newtonian fluid provided in the first aspect of the embodiment of the present application. Figure 1 In the method, the performance detection method of the non-Newtonian fluid includes:

[0054] S1. When the surface of the non-Newtonian fluid is in a stationary state, obtain a reference height of the surface of the non-Newtonian fluid.

[0055] A non-Newtonian fluid is a fluid that does not satisfy Newton's experimental law of viscosity, that is, a fluid whose shear stress and shear strain rate are not linearly related. In the present embodiment, the non-Newtonian fluid is concrete. In other embodiments, the non-Newtonian fluid can also be oil, mud, asphalt, petroleum, or other substances, and the specific configuration can be determined by those skilled in the art based on actual needs.

[0056] The embodiment of the present application uses a laser ranging sensor to obtain the reference height.

[0057] S2. When the surface of the non-Newtonian fluid is in a non-stationary state, obtain a height dataset of the surface of the non-Newtonian fluid based on a time series.

[0058] When the surface of the non-Newtonian fluid is in a non-stationary state, that is, the non-Newtonian fluid is affected by an external force (such as stirring, vibration, impact, etc.), the surface of the non-Newtonian fluid is deformed.

[0059] When the frequency of the external force is 20 to 100 Hz, the timing curve reflects the macroscopic rheological properties and the response of the overall structure; when the frequency of the external force is 100 to 1000 Hz, the timing curve reflects the microscopic structural changes and the interaction between particles; when the frequency of the external force is greater than 1000 Hz, the timing curve reflects the molecular-level structural characteristics and the movement of molecular chains.

[0060] The height dataset is used to reflect the height change of the surface of the non-Newtonian fluid during the time series.

[0061] The embodiment of the present application uses the laser ranging sensor to obtain the height data set.

[0062] S3. Construct a time series curve based on the reference altitude and the altitude data set.

[0063] The S3 steps include:

[0064] S31. Calculate the difference between the reference height and each height data in the height data set to obtain target height change data corresponding to each height data.

[0065] Specifically, the calculation formula for the target height change data corresponding to the target height data is: ;in, is target height change data corresponding to the target height data; is the target height data, that is, the height data set Altitude data corresponding to the moment; is the reference height.

[0066] S32: Construct the timing curve based on all target height change data.

[0067] Specifically, all target height change data are connected in time series to form the time series curve.

[0068] The timing curve can be used to characterize the thixotropy of the non-Newtonian fluid. In some embodiments, when constructing the timing curve, a data cleaning operation is performed on all target height change data.

[0069] S4. Perform feature extraction processing on the timing curve to obtain at least one feature parameter.

[0070] The at least one characteristic parameter includes a maximum amplitude, a decay rate, an oscillation frequency, and a stabilization time. In other embodiments, the at least one characteristic parameter may also include temperature, an external excitation frequency, or other content, which can be specifically set by a person skilled in the art according to actual needs. The external excitation frequency refers to the frequency at which an external force stirs, vibrates, or impacts the non-Newtonian fluid.

[0071] Furthermore, step S4 includes: taking the absolute values ​​of all target height change data on the timing curve, and taking the largest value among all absolute values ​​as the maximum amplitude; using the fit function (also known as the fitting function) of MATLAB (MATLAB is a commercial mathematical software) to estimate the timing curve to obtain the attenuation rate; converting the timing curve into a frequency domain signal, and obtaining the power spectrum density from the frequency domain signal; finding the frequency corresponding to the maximum peak in the power spectrum density to obtain the oscillation frequency; setting a time window and a step size; dividing the timing curve into multiple segments of curves to be processed based on the time window and the step size; when the difference in the means corresponding to two adjacent segments of curves to be processed is less than a first preset threshold or the difference in the variances corresponding to two adjacent segments of curves to be processed is less than a second preset threshold, the time points corresponding to the two adjacent segments of curves to be processed are taken as the stabilization time.

[0072] The calculation formula for the step of taking the absolute values ​​of all target height change data on the timing curve and taking the largest value among all absolute values ​​as the maximum amplitude can be expressed as: ; is the maximum amplitude, The function is a mathematical function that returns the maximum value in a set of values.

[0073] The maximum amplitude is used to reflect the maximum response degree of the fluid to the external force.

[0074] The time series curve is estimated using the fit function of MATLAB, and the calculation formula for the step of obtaining the attenuation rate is: ;in, is Equal to the amplitude of the timing curve at time 0; is the base of natural logarithms; is the decay rate; is a time variable; is the angular frequency; It is the initial phase.

[0075] The attenuation rate characterizes the structural recovery and energy dissipation characteristics of the fluid, is used to reflect the thixotropy and viscoelasticity of the fluid, and can also be used to characterize the rheological properties of the non-Newtonian fluid.

[0076] The calculation formula of the oscillation frequency is: ;in, is the oscillation frequency; is a function that finds the maximum peak in the power spectral density; is the power spectral density function; is the frequency domain signal.

[0077] The oscillation frequency reflects the inherent vibration characteristics of the internal structure of the fluid, and the oscillation frequency is related to the elastic modulus and density of the fluid. It should be understood that the fluids described herein all refer to the non-Newtonian fluids.

[0078] The calculation formula for the case where the mean difference between two adjacent segments of the curve to be processed is less than the first preset threshold is: ;in, is the arithmetic mean function; It is The curve to be processed; It is The curve to be processed; is the first preset threshold.

[0079] The calculation formula for whether the difference in variances between two adjacent segments of the curve to be processed is less than the second preset threshold is: ; is the variance function; is the second preset threshold.

[0080] The stabilization time is the time required for the fluid to recover from a disturbed state to a stable state, and is used to reflect the structural reconstruction ability and thixotropic recovery characteristics of the fluid.

[0081] The specific values ​​of the time window, the step size, the first preset threshold and the second preset threshold can be set by those skilled in the art according to actual needs.

[0082] S5. Use a preset algorithm to detect and process the at least one characteristic parameter to obtain performance parameters of the non-Newtonian fluid.

[0083] Through the above-described implementation, the embodiments of the present application utilize a reference height and height dataset of the non-Newtonian fluid surface, which are directly associated with the critical state of the non-Newtonian fluid, to detect performance parameters, thereby improving the accuracy of detecting the performance parameters of the non-Newtonian fluid. Furthermore, the reference height and height dataset of the non-Newtonian fluid surface are not subject to the influence of the structure, material properties, and fluid coupling effects of the additional equipment, thereby improving the accuracy of detecting the performance parameters of the non-Newtonian fluid.

[0084] The performance detection method for non-Newtonian fluids can detect performance parameters without sampling the non-Newtonian fluids, thereby simplifying the detection steps of traditional performance detection methods for non-Newtonian fluids.

[0085] The performance detection method of non-Newtonian fluids can avoid the interference caused by the contact between the probe and the fluid in traditional methods, eliminate the influence of equipment material properties and fluid coupling effects, is suitable for harsh environments such as high temperature and corrosion, and can perform real-time detection to support the integration of automated control systems.

[0086] Furthermore, the step of S5 includes:

[0087] S51 . Detect and process the at least one characteristic parameter using the yield stress model of the preset algorithm to obtain a yield stress parameter of the performance parameter.

[0088] The calculation formula of the yield stress model is:

[0089] ;

[0090] in, is the yield stress parameter of the yield stress model; is the first fitting coefficient; is the second fitting coefficient; is the maximum amplitude; is the third fitting coefficient; is the decay rate; is the fourth fitting coefficient; is the oscillation frequency; is the fifth fitting coefficient; is the stabilization time; is the sixth fitting coefficient; is the seventh fitting coefficient; is the eighth fitting coefficient; is the ninth fitting coefficient; is the tenth fitting coefficient; The yield stress parameter reflects the minimum shear stress required for the fluid to begin flowing and is used in pumping pressure design and pipeline transportation calculation scenarios.

[0091] More specifically, the calculation formula of the yield stress model is:

[0092] .

[0093] In some embodiments, the yield stress model is obtained by fitting and improving the first base model, the second base model, the third base model, and the fourth base model.

[0094] The calculation formula of the first basic model is:

[0095] ;

[0096] in, is the yield stress parameter of the first basic model; is the twelfth fitting coefficient; is the thirteenth fitting coefficient; is the fourteenth fitting coefficient; is the fifteenth fitting coefficient; is the sixteenth fitting coefficient; is the seventeenth fitting coefficient.

[0097] The calculation formula of the second basic model is:

[0098] ;

[0099] in, is the yield stress parameter of the second basis model; is the eighteenth fitting coefficient; is the nineteenth fitting coefficient; is the twentieth fitting coefficient; is the twenty-first fitting coefficient; is the twenty-second fitting coefficient; is the twenty-third fitting coefficient.

[0100] The calculation formula of the third basic model is:

[0101] ;

[0102] in, is the yield stress parameter of the third basic model; is the twenty-fourth fitting coefficient; is the twenty-fifth fitting coefficient; is the twenty-sixth fitting coefficient; is the twenty-seventh fitting coefficient; is the twenty-eighth fitting coefficient; is the twenty-ninth fitting coefficient.

[0103] The calculation formula of the fourth basic model is:

[0104] ;

[0105] in, is the quadratic correction coefficient of the fourth basic model; is the 30th fitting coefficient; is the thirty-first fitting coefficient; is the thirty-second fitting coefficient; is the thirty-third fitting coefficient; is the thirty-fourth fitting coefficient; is the thirty-fifth fitting coefficient.

[0106] Furthermore, the first basic model may be a Bingham model (also known as the Bingham model); the second basic model may be a Herschel-Bulkley model (also known as the Herschel-Bulkley model); the third basic model may be a Modified Bingham model (also known as the Modified Bingham model); and the fourth basic model may be a quadratically modified Modified Bingham model.

[0107] In some embodiments, the yield stress model may also be a trained multilayer perceptron, a long short-term memory network, or a support vector machine. In the case where the yield stress model is a trained long short-term memory network, the time step of the long short-term memory network is 50.

[0108] In other embodiments, the yield stress model may also be other deep learning networks, which may be specifically configured by those skilled in the art according to actual needs.

[0109] In some embodiments, step S51 may include:

[0110] S511 . Perform feature extraction processing on the at least one characteristic parameter using the first hidden layer of the yield stress model to obtain a first characteristic vector.

[0111] The number of neurons in the first hidden layer is set to 64.

[0112] The specific implementation of step S511 is that the first hidden layer uses the first weight matrix and the first bias to perform linear transformation processing on the at least one feature parameter, thereby obtaining the first feature vector.

[0113] Specifically, the calculation formula of step S511 may be: ; is the first eigenvector; is the at least one characteristic parameter, ; is the first weight matrix; is the first bias.

[0114] In other implementations, the calculation formula of step S511 may also be other formulas, which can be specifically set by those skilled in the art according to actual needs.

[0115] S512: Use the first target activation function of the yield stress model to perform nonlinear transformation processing on the first eigenvector to obtain a second eigenvector.

[0116] The first target activation function is a ReLU activation function. In other embodiments, the first target activation function is not limited to the ReLU activation function, and can be specifically set by those skilled in the art according to actual needs.

[0117] The calculation formula of step S512 is: ; is the second eigenvector.

[0118] S513: Use the second hidden layer of the yield stress model to perform feature extraction processing on the second eigenvector to obtain a third eigenvector.

[0119] The number of neurons in the second hidden layer is set to 32.

[0120] The specific implementation of step S513 is that the second hidden layer uses the second weight matrix and the second bias to perform linear transformation processing on the second eigenvector, so as to obtain the third eigenvector.

[0121] The calculation formula of step S513 can be: ; is the third eigenvector; is the second weight matrix; is the second bias.

[0122] S514: Use the first target activation function to perform nonlinear transformation processing on the third eigenvector to obtain a fourth eigenvector.

[0123] The calculation formula of step S514 is: ; is the fourth eigenvector.

[0124] S515 : Using the third hidden layer of the yield stress model to perform feature extraction processing on the fourth eigenvector to obtain a fifth eigenvector.

[0125] The number of neurons in the third hidden layer is set to 16.

[0126] The specific implementation of step S515 is that the third hidden layer uses the third weight matrix and the third bias to perform linear transformation processing on the fourth eigenvector, so as to obtain the fifth eigenvector.

[0127] The calculation formula of step S515 can be: ; is the fifth eigenvector; is the third weight matrix; is the third bias.

[0128] S516. Use the first target activation function to perform nonlinear transformation processing on the fifth eigenvector to obtain the yield stress parameter.

[0129] The calculation formula for S516 is: ; is the yield stress parameter.

[0130] S52: Using the plastic viscosity model of the preset algorithm to detect and process the at least one characteristic parameter to obtain a plastic viscosity parameter of the performance parameter.

[0131] The calculation formula of the plastic viscosity model is:

[0132] ;

[0133] in, is the plastic viscosity parameter output by the plastic viscosity model; is the thirty-sixth fitting coefficient; is the thirty-seventh fitting coefficient; is the thirty-eighth fitting coefficient; is the thirty-ninth fitting coefficient; is the 40th fitting coefficient.

[0134] In some embodiments, the plastic viscosity model may also be a trained multilayer perceptron, long short-term memory network, or support vector machine. In other embodiments, the plastic viscosity model may also be other deep learning networks, which may be specifically configured by those skilled in the art according to actual needs.

[0135] Furthermore, the calculation formula of step S52 may also be: ; ;in, is the sixth eigenvector; is the fourth weight matrix; is the first learnable parameter; is the second learnable parameter; is the fourth bias; is the plastic viscosity parameter.

[0136] In some embodiments, the plastic viscosity model can also introduce residual connections to improve training stability. Specifically, the residual connection calculation formula is ; is the output of the residual connection, which is used to alleviate the training problem of deep networks; yes projection.

[0137] In some embodiments, the performance detection method of the non-Newtonian fluid also includes a training step of the yield stress model, and the training step of the yield stress model includes: obtaining a historical data set and a first real label set, wherein each first real label in the first real label set has a one-to-one correspondence with one of the historical data in the historical data set; using the original yield stress model to detect and process the target historical data to obtain a first prediction result, and the target historical data is any historical data in the historical data set; constructing a first loss function based on the first prediction result and the first real label corresponding to the target historical data; adjusting the first model parameter of the original yield stress model based on the first loss function and the first preset constraint item to form the yield stress model.

[0138] Each historical data in the historical data set includes a historical maximum amplitude, a historical decay rate, a historical oscillation frequency, and a historical stabilization time. In other embodiments, each historical data may also include environmental parameters, ultrasonic frequency, or other content.

[0139] The historical maximum amplitude ranges from 0.1 to 5.0 mm, the historical decay rate ranges from 0.05 to 2.0 Hz, the historical oscillation frequency ranges from 0.1 to 50 Hz, the environmental parameter ranges from 5 to 40°C, and the ultrasonic frequency ranges from 20 to 10,000 Hz. Based on the above data, the first prediction result ranges from 50 to 500 Pascals.

[0140] The first loss function is a mean square error loss function.

[0141] The first preset constraint item is that the first prediction result is greater than or equal to 0, and the weight coefficient is equal to 0.2.

[0142] The first loss function and the first preset constraint item can be set according to actual needs.

[0143] In some embodiments, after obtaining the historical data set, the yield stress model training step further includes: performing data cleaning, Z-score (the full name of Z-score in English is standardscore, meaning standard score) normalization and data enhancement processing on the historical data set.

[0144] The data cleaning specifically involves removing abnormal data with an amplitude greater than 3σ or a sampling rate less than 1 kHz. The Z-score normalization has a mean of 0 and a standard deviation of 1, ensuring that all features have the same scale. Data augmentation involves adding Gaussian noise (σ = 0.1 mm) and a time offset (±50 milliseconds). In some embodiments, the yield stress model training step further includes evaluating the yield stress model using a target evaluation method to obtain a first evaluation result; and adjusting second model parameters of the yield stress model based on the first evaluation result.

[0145] The target evaluation method is five-fold cross validation. In other embodiments, other evaluation methods can also be used, which can be set by those skilled in the art according to actual needs.

[0146] During the evaluation process of the target evaluation method, the early stopping threshold is set to 10 rounds and the batch size is 32.

[0147] In some embodiments, the yield stress model training step further comprises: optimizing the yield stress model using Dropout and L2 regularization.

[0148] The value of Dropout is preferably 0.2, and the regularization coefficient of L2 regularization is 0.001. In other embodiments, the value of Dropout and the regularization coefficient can be set by those skilled in the art according to actual needs.

[0149] In some embodiments, the yield stress model training step further comprises optimizing the yield stress model using a constraint condition, wherein the constraint condition is a prediction result, that is, an output result of the yield stress model is greater than or equal to 0, and a weight coefficient is 0.2.

[0150] In some embodiments, the yield stress model training step further includes: hyperparameter tuning. The hyperparameters include a learning rate, a tree depth, and a kernel function parameter, wherein the learning rate is 1e-4 to 1e-2, the tree depth is 3 to 10, and the kernel function parameter is 0.1 to 10.

[0151] During the training of the yield stress model, in terms of feature selection, characteristic parameters with a SHAP value greater than 0.1 are retained. The SHAP value is used to rank the importance of the at least one characteristic parameter.

[0152] In the verification stage, the test set accounts for 20% and the prediction error threshold of the first prediction result is ±15Pa; in the physical inspection stage, the slight change in the first prediction result and the slight change in the historical maximum amplitude are less than 0.

[0153] In some embodiments, the performance detection method of the non-Newtonian fluid also includes a training step of the plastic viscosity model, and the training step of the plastic viscosity model includes: obtaining a second real label set, wherein each second real label in the second real label set has a one-to-one correspondence with one of the historical data in the historical data set; using the original plastic viscosity model to detect and process the target historical data to obtain a second prediction result; constructing a second loss function based on the second prediction result and the second real label corresponding to the target historical data; adjusting the third model parameter of the original plastic viscosity model based on the second loss function and the second preset constraint item to form the plastic viscosity model.

[0154] The second loss function is a mean square error loss function.

[0155] The second preset constraint item is that the second prediction result is greater than or equal to 0, and the weight coefficient is equal to 0.2.

[0156] In some embodiments, after obtaining the historical data set, the step of training the plastic viscosity model further includes: performing data cleaning, Z-score normalization, and data enhancement processing on the historical data set.

[0157] In some embodiments, the present invention can also adjust the fifth model parameter of the plastic viscosity model and the sixth model parameter of the yield stress model by using a third loss function. The calculation formula of the third loss function is as follows:

[0158] ;

[0159] in, is the fifth weight coefficient; is the static parameter loss; is the sixth weight coefficient; is the dynamic parameter loss; is the seventh weight coefficient; is the time-dependent parameter loss; is the L2 regularization term; is the regularization coefficient. The static parameters include yield stress parameter, static viscosity, flow index, and thixotropy index. The dynamic parameters include storage modulus, loss modulus, dynamic viscosity, and loss factor. The time-dependent parameters include structural recovery time, shear thinning index, and thixotropic loop area.

[0160] In some embodiments, the training step of the plastic viscosity model further includes: evaluating the plastic viscosity model using the target evaluation method to obtain a second evaluation result; and adjusting a fourth model parameter of the plastic viscosity model based on the second evaluation result.

[0161] In some embodiments, the step of training the plastic viscosity model further comprises: optimizing the plastic viscosity model using Dropout and L2 regularization.

[0162] During the verification phase, the prediction error threshold of the second prediction result is ±10%.

[0163] The yield stress model and the plastic viscosity model can be understood as machine learning models. The training data of the machine learning models can be obtained by a standard rheometer to ensure the accuracy and reliability of the data.

[0164] The standard rheometer can be a rotational rheometer, an oscillatory rheometer, a capillary rheometer, or other rheometers. The rotational rheometer uses a stress sweep mode to measure the yield stress, a shear rate sweep to measure the viscosity curve, and a time sweep and cyclic shear test to test the thixotropy; the oscillatory rheometer uses a frequency range of 0.01 to 100 Hz to measure the storage modulus and loss modulus, a strain sweep to determine the linear viscoelastic region, and a temperature sweep to determine the rheological properties at different temperatures; the capillary rheometer uses a shear rate range of 1000 to 10,000 per second to determine the high shear viscosity and a power law model parameter to determine the shear thinning behavior.

[0165] In some embodiments, the machine learning model may further include a shared feature extraction layer, a first branch network, a second branch network, and a third branch network.

[0166] The shared feature extraction layer is a ReLU activation function. The input of the shared feature extraction layer is the result of the formula W_shared×X+b_shared, where W_shared is a shared weight matrix, X is the at least one feature parameter, and b_shared is a shared bias. The output of the shared feature extraction layer is the shared feature. The first branch network includes multiple fully connected layers and multiple activation functions. The input of the first branch network is the shared feature, and the output is the yield stress parameter, static viscosity, and flow index. The second branch network includes multiple fully connected layers and multiple activation functions, and each output parameter corresponds to a fully connected layer and an activation function. The input of the second branch network is the shared feature, and the output is the complex viscosity, storage modulus, and loss modulus. Each output parameter corresponds to a fully connected layer and an activation function. The third branch network includes multiple fully connected layers and multiple activation functions. The input of the third branch network is the shared feature, and the output is the structural recovery time and thixotropic index. Each output parameter corresponds to a fully connected layer and an activation function.

[0167] The machine learning model can also introduce an attention mechanism to strengthen the shared features.

[0168] In some embodiments, after step S5, the method for detecting the performance of a non-Newtonian fluid further includes obtaining temperature data of the non-Newtonian fluid and performing temperature compensation based on the temperature data to improve the detection accuracy of the embodiment of the present application.

[0169] Furthermore, the temperature data may be temperature compensated using a temperature correction model.

[0170] The calculation formula of the temperature correction model is:

[0171] ;

[0172] in, is the yield stress parameter at the test temperature; is the yield stress parameter at the reference temperature; is a natural exponential function; is the activation energy, which can be likened to the “energy” required to produce a temperature-dependent change in yield stress; is the gas constant, which is 8.314 J / mol·K (J / mol·K is the unit, joule / mole Kelvin); is the test temperature; is a reference temperature; the test temperature may be data used to perform temperature compensation on the temperature data.

[0173] Furthermore, The value is 25000 J / mol (J / mol is an energy unit, which means the energy per mole of substance). The value is 293.15K (K is the Kelvin unit).

[0174] In other embodiments, the calculation formula of the temperature correction model may also be:

[0175] ;

[0176] in, The temperature data is the data after temperature compensation is performed on the temperature data; is the temperature correction function.

[0177] In some embodiments, after step S5, the performance detection method of the non-Newtonian fluid further includes analyzing the response characteristics of the surface of the non-Newtonian fluid at different frequencies and amplitudes, wherein the response characteristics include amplitude response ratio, phase difference, and harmonic components; performing spectral analysis on the amplitude response ratio, the phase difference, and the harmonic components to generate a three-dimensional spectrum containing full-band response characteristics; and using the preset algorithm to detect and process the three-dimensional spectrum to obtain the comprehensive performance characteristics of the non-Newtonian fluid.

[0178] The amplitude response ratio is used to detect the energy dissipation characteristics of the non-Newtonian fluid and reflect the viscoelastic parameters of the non-Newtonian fluid; the phase difference is used to quantify the rheological hysteresis effect of the non-Newtonian fluid and characterize the thixotropy and structural recovery ability of the non-Newtonian fluid; the harmonic component is used to identify the nonlinear rheological characteristics of the non-Newtonian fluid and reflect the internal structural changes of the material and the interaction between particles.

[0179] In the application of using the preset algorithm to detect and process the three-dimensional spectrum to obtain the comprehensive performance characteristics of the non-Newtonian fluid, the learning process of the preset algorithm includes: in the data preparation stage, collecting the characteristic spectrum data of the non-Newtonian fluid with different ratios under the sweep frequency of 20 Hz to 10 kHz, and synchronously obtaining the performance parameters measured by the standard rheometer; in the feature engineering stage, extracting the key parameters in the characteristic spectrum data, including the resonance frequency, half-peak width, phase turning point and harmonic distortion rate, the resonance frequency reflects the inherent characteristics of the fluid, the half-peak width characterizes the damping characteristics, the phase turning point reflects the critical point of the structural transition, and the harmonic distortion rate. The variability is a quantitative indicator reflecting the degree of nonlinearity; in the model construction stage, the input layer of the deep convolutional neural network (DCNN) in the preset algorithm is used to receive the two-dimensional spectrum matrix (frequency × amplitude or frequency × phase), and the output layer predicts the rheological parameters (i.e., yield stress parameters and plastic viscosity parameters, etc.). In the model optimization stage, the key frequency band features (50 to 500 Hz rheological sensitive area) are enhanced through the attention mechanism, and the hyperparameters are adjusted by Bayesian optimization to optimize the feature extraction effect; in the verification stage, the five-fold cross-validation is used to evaluate the generalization ability of the model in the mapping of characteristic spectrum to performance parameters, and finally achieve Detection accuracy greater than 0.95.

[0180] In other implementations, the response characteristics are not limited to the amplitude response ratio, the phase difference, and the harmonic component, and can be specifically set by those skilled in the art according to actual needs.

[0181] In some embodiments, the performance parameter also includes a wet billet strength parameter. For a non-Newtonian fluid during the solidification process, the change in the amplitude of the response of the surface height of the non-Newtonian fluid to an external force can be used to characterize the wet billet strength of the non-Newtonian fluid. Embodiments of the present application can assess the wet billet strength parameter in real time.

[0182] In some embodiments, the performance parameter further includes a consistency coefficient. The calculation formula of the consistency coefficient is:

[0183] ;

[0184] in, is the consistency coefficient; is the 41st fitting coefficient; is the 42nd fitting coefficient; is the 43rd fitting coefficient; is the 44th fitting coefficient; is the 45th fitting coefficient.

[0185] In some embodiments, the performance parameter further includes a flow index. The flow index is calculated as follows:

[0186] ;

[0187] in, is the flow index; is the 46th fitting coefficient; is the 47th fitting coefficient; The flow index is used to describe the degree to which a fluid deviates from a Newtonian fluid and is used in application scenarios such as quantifying non-Newtonian properties and guiding formulation optimization.

[0188] In some embodiments, the performance parameter further includes static viscosity. The static viscosity is calculated as follows:

[0189] ;

[0190] in, is the static viscosity; is the infinite shear viscosity; is the zero shear viscosity; is the time constant. The static viscosity reflects the apparent viscosity at low shear rates and is used in stirring power calculations and fluidity assessments.

[0191] In some embodiments, the performance parameter further includes dynamic viscosity, and the dynamic viscosity is calculated as follows:

[0192] ;

[0193] ;

[0194] in, is the complex viscosity; is the dynamic viscosity, also known as the real part of viscosity; It is the imaginary unit, used to distinguish the real part from the imaginary part; is the loss viscosity, also called the imaginary part of viscosity; The complex viscosity is used in processing performance prediction and process parameter optimization scenarios.

[0195] In some embodiments, the performance parameters further include storage modulus and loss modulus. The calculation formulas for the storage modulus and loss modulus are:

[0196] ;

[0197] ;

[0198] in, is the storage modulus; is the equilibrium modulus; It’s relaxation time; The storage modulus characterizes the elastic properties of the material and is used in material stiffness assessment and structural strength prediction scenarios. The loss modulus characterizes the viscous properties of the material and is used in energy dissipation analysis and damping property assessment scenarios. In other embodiments, the present application also includes a loss factor for characterizing the viscoelastic ratio. The loss factor is .

[0199] In other embodiments, the calculation formulas for the storage modulus and loss modulus may also be:

[0200] ;

[0201] ;

[0202] in, It is The equilibrium modulus of the relaxation process; It is The relaxation time of the relaxation process.

[0203] In some embodiments, the performance parameter further includes a structural recovery time. The calculation formula for the structural recovery time is:

[0204] ;

[0205] in, is the structural recovery time; is the pre-factor; is the activation energy of recovery. The structural recovery time is the time required for the fluid structure to fully recover, and is used in scenarios such as determining the resting time and designing process intervals.

[0206] In some embodiments, the performance parameter further includes a thixotropic index. The calculation formula of the thixotropic index is: ;

[0207] in, is the thixotropic index; is the thixotropic constant; is the combined index of yield stress and viscosity; The frequency and structure recovery index characterizes the fluid structure destruction and reconstruction ability.

[0208] In some embodiments, the present invention can also simultaneously optimize the fitting accuracy of multiple rheological parameters. The specific calculation formula is:

[0209] ;

[0210] in, It is the fitting data for optimizing multiple rheological parameters simultaneously; is the first weight coefficient; is the mean square error function; is the second weight coefficient; is the third weight coefficient; is the fourth weight coefficient.

[0211] In some embodiments, the performance parameters also include a shear thinning index and a thixotropic loop area. The shear thinning index describes the degree of shear thinning and is used in flow behavior prediction and transport performance assessment. The thixotropic loop area represents the area enclosed by the ascending and descending curves and is used in stirring effect assessment and uniformity determination.

[0212] For the quality assessment after fitting, the coefficient of determination is greater than 0.85; the root mean square error (RMSE) is less than 15%; the mean absolute error (MAE) is less than 12%; the significance level of the F test is α=0.05.

[0213] In practical applications, the present embodiment employs a single impact response measurement method to apply a single impact or vibration to the surface of the non-Newtonian fluid to obtain the timing curve. More specifically, a mechanical impactor or short-pulse vibrator is used to apply a quantitative impact to the surface of the non-Newtonian fluid; a laser ranging sensor is used to continuously record the height recovery process of the surface; and the rheological parameters of the non-Newtonian fluid are calculated by analyzing the timing curve of the height recovery process of the surface.

[0214] In other application scenarios, the embodiments of the present application can also use a periodic vibration response measurement method to apply periodic vibrations of different frequencies to the surface of the non-Newtonian fluid to record the amplitude response and phase difference of the surface height. The periodic vibration response measurement method is mainly suitable for evaluating the thixotropy and frequency dependence of the non-Newtonian fluid. The specific implementation includes: using a vibrator to apply fixed-frequency vibrations to the non-Newtonian fluid; using the laser ranging sensor to record the vibration response of the surface height; based on the vibration response, analyzing the amplitude ratio and phase difference of the surface response at different vibration frequencies, and constructing an amplitude-frequency relationship curve; based on the amplitude-frequency relationship curve, evaluating the thixotropy and non-Newtonian characteristics of the non-Newtonian fluid.

[0215] In other application scenarios, the embodiments of the present application can also use an ultrasonic sweep frequency response measurement method to perform full-spectrum characteristic analysis on the non-Newtonian fluid. The specific implementation includes: using an ultrasonic generator to generate a sweep frequency vibration of 20 Hz to 10 kHz according to a preset program; using a laser ranging sensor to record the height response data of the surface in real time; based on the height response data, analyzing the amplitude, phase and harmonic components of the response at different frequencies; establishing a frequency-response characteristic spectrum based on the amplitude, phase and harmonic components of the response at different frequencies to evaluate the comprehensive rheological properties of the non-Newtonian fluid.

[0216] In other application scenarios, the embodiments of the present application can also use the natural disturbance measurement method of the stirring process to realize real-time detection of the rheological properties of the non-Newtonian fluid. The specific implementation includes: installing the laser ranging sensor above the stirring device and aligning it with a specific measurement area on the surface of the non-Newtonian fluid; continuously recording the height change of the surface during the stirring process; identifying the characteristic waveform generated when the blade of the stirring device passes through the specific measurement area; analyzing the attenuation characteristics of the characteristic waveform to evaluate the uniformity and rheological properties of the non-Newtonian fluid; when the attenuation characteristics reach the preset conditions, the stirring uniformity of the non-Newtonian fluid is determined, and the next step can be performed.

[0217] The preset conditions can be set by those skilled in the art according to actual needs.

[0218] When the non-Newtonian fluid is concrete, the method for detecting the setting state of shotcrete is as follows: During tunnel shotcrete construction, a laser ranging sensor is installed on the shotcrete equipment to monitor the shotcrete layer. By applying a measured impact or vibration to the concrete surface and analyzing the surface deformation characteristics, the setting state of the concrete is assessed. More specifically, the laser ranging sensor and a small vibrator are mounted on the same bracket and aimed at the shotcrete layer. The vibrator then generates a measured vibration to disturb the concrete surface. The laser ranging sensor then records the surface height response. By analyzing the characteristics of the response curve (such as amplitude and attenuation rate), the setting state of the concrete is assessed. Based on the setting state, a decision is made as to whether the next shotcrete layer or other construction operations can be performed.

[0219] In the case where the non-Newtonian fluid is concrete, the method for testing the rheological properties of concrete during pumping is as follows: a transparent observation window is installed at an appropriate location in the concrete pumping pipeline to ensure that laser light can penetrate the transparent observation window and illuminate the concrete surface; a laser ranging sensor is used to detect the concrete surface within the pipeline through the transparent observation window. More specifically, the flow characteristics of the concrete under pumping pressure are analyzed to evaluate pumping performance; the laser ranging sensor is aligned with the transparent observation window to measure the height change of the concrete surface; during the pumping process, the fluctuation characteristics of the concrete surface are recorded and a fluctuation curve is constructed based on the fluctuation characteristics; the rheological properties and pumping suitability of the concrete are evaluated by analyzing the frequency and amplitude characteristics of the fluctuation curve; when abnormal fluctuation characteristics are detected, a timely warning is issued to prevent problems such as pipeline blockage.

[0220] In the case where the non-Newtonian fluid is concrete, during the mixing process of the non-Newtonian fluid, it is necessary to identify the characteristic waveform of the blade passing through the measurement area, analyze the attenuation characteristics to evaluate the uniformity, determine the mixing completion in real time, analyze the fluctuation characteristics in the transparent observation window, evaluate the pumping adaptability based on the frequency and amplitude characteristics, and provide real-time abnormal fluctuation warnings to prevent pipeline blockage.

[0221] In the case where the non-Newtonian fluid is concrete, the method for testing the wet-sheet strength of 3D-printed concrete is as follows: a laser rangefinder is installed on the 3D concrete printing equipment to test the newly printed concrete layer. By analyzing the concrete surface's response to vibration, its wet-sheet strength is assessed to ensure the safety of the printed layer. More specifically, a laser rangefinder and a small vibrator are installed next to the 3D print head. After a layer is printed, a constant vibration is applied to the printed layer. The laser rangefinder records the surface height change response. By analyzing the characteristics of the response curve, the wet-sheet strength data of the concrete is assessed. Based on the wet-sheet strength data, the printing interval or the printing parameters of the next layer are automatically adjusted.

[0222] The non-Newtonian fluid performance testing method provided in the embodiments of this application can be integrated into existing concrete production and construction equipment to achieve automated detection and control. Specifically, the non-Newtonian fluid performance testing method can be integrated into a mixing plant control system, a pumping pressure control system, a printing control system, or other automated control systems to detect the performance parameters of the non-Newtonian fluid and automatically adjust the mix ratio or process parameters of the non-Newtonian fluid.

[0223] In some embodiments, the performance detection method of the non-Newtonian fluid may also involve the content of quantum distance measurement. Specifically, a graphene quantum dot laser emitter, an integrated environmental noise quantum sensing unit and a nano-level vibration compensation algorithm are designed. Among them, the graphene quantum dot laser emitter is used to measure the distance between the non-Newtonian fluid surface and the sensor, the integrated environmental noise quantum sensing unit is used to detect the environmental noise generated when the non-Newtonian fluid is stirred by an external force, and the nano-level vibration compensation algorithm is used to eliminate or suppress the influence of vibration on the measurement of the distance between the non-Newtonian fluid surface and the sensor.

[0224] refer to Figure 2 The figure shows a principle block diagram of a performance detection device for non-Newtonian fluids provided in the second aspect of the embodiment of the present application. Figure 2 In the embodiment, the performance detection device 100 of the non-Newtonian fluid comprises:

[0225] A reference height acquisition module 101 is configured to acquire a reference height of a surface of a non-Newtonian fluid when the surface of the non-Newtonian fluid is in a stationary state;

[0226] A height dataset acquisition module 102 is configured to acquire a height dataset of the surface of the non-Newtonian fluid based on a time sequence when the surface of the non-Newtonian fluid is in a non-stationary state;

[0227] A timing curve construction module 103 is configured to construct a timing curve based on the reference height and the height data set;

[0228] A feature extraction module 104 is configured to perform feature extraction processing on the timing curve to obtain at least one feature parameter;

[0229] The performance parameter obtaining module 105 is configured to detect and process the at least one characteristic parameter using a preset algorithm to obtain the performance parameter of the non-Newtonian fluid.

[0230] refer to Figure 3 The figure shows a principle block diagram of a performance detection system for non-Newtonian fluids provided in the third aspect of the embodiment of the present application. Figure 3 In the embodiment, the non-Newtonian fluid performance detection system 200 includes a laser ranging module 201, an external force application module 202, a data acquisition module 203 and a terminal device 204, and the terminal device 204 includes a processor and a memory. The processor is communicatively connected with the laser ranging module 201, the external force application module 202 and the data acquisition module 203, the memory is used to store computer programs, and the processor is used to call and run the computer programs stored in the memory to execute the steps of the non-Newtonian fluid performance detection method provided in the first aspect of the above-mentioned embodiment of the present application.

[0231] The processor is further configured to control the working states of the laser ranging module 201 and the external force application module 202 .

[0232] The memory is further used to store a historical data set, a first model parameter, a second model parameter, a third model parameter and a fourth model parameter.

[0233] In some embodiments, the laser ranging module 201 includes a laser ranging sensor, a fixed bracket and a protective cover, the fixed bracket is connected to the laser ranging sensor, and the laser ranging sensor is arranged in the protective cover; the laser ranging sensor is used to emit laser light to the surface of the non-Newtonian fluid and receive the reflected laser light to calculate the distance between the surface of the non-Newtonian fluid and the laser ranging sensor; the fixed bracket is used to adjust the position of the laser ranging sensor and the irradiation angle of the laser of the laser ranging sensor; the protective cover is used to prevent dust and spatters to prevent the measurement accuracy of the laser ranging sensor from being affected by dust and spatters.

[0234] The laser distance measuring sensor is arranged above the surface of the non-Newtonian fluid, and the laser is vertically irradiated onto a specific point on the surface of the non-Newtonian fluid.

[0235] The measurement accuracy of the laser distance sensor is better than 0.1 mm (mm is millimeter).

[0236] The laser ranging sensor is also used to record in real time the response curve (also known as the time series curve) of the surface of the non-Newtonian fluid under different ultrasonic parameters.

[0237] The number of the laser ranging sensors may be one or more, so as to obtain more comprehensive surface deformation information.

[0238] In some embodiments, the laser ranging module 201 further includes a pressure sensor to establish a more comprehensive detection system.

[0239] In some embodiments, the external force application module 202 includes a vibrator, an impactor and / or an ultrasonic generator, wherein the vibrator is used to apply vibration to the non-Newtonian fluid at different frequencies (0.1 to 50 Hz) and amplitudes; the impactor is used to apply quantitative impact to the surface of the non-Newtonian fluid; and the ultrasonic generator is used to generate ultrasonic vibrations of different frequencies (20 Hz to 10 kHz) and amplitudes to apply vibration to the non-Newtonian fluid.

[0240] The processor can be communicatively connected to the vibrator, the impactor and / or the ultrasonic generator. The processor can send an apply vibration command to the vibrator and / or the ultrasonic generator. After receiving the apply vibration command, the vibrator and / or the ultrasonic generator starts to operate and generate vibration, so that the non-Newtonian fluid is vibrated as a whole and the surface of the non-Newtonian fluid is deformed; the processor can send an apply quantitative impact command to the impactor. After receiving the apply quantitative impact command, the impactor starts to operate and generates shaking, so that the surface of the non-Newtonian fluid is deformed.

[0241] The ultrasonic generator is usually installed on the bottom or side wall of the container of the non-Newtonian fluid. The surface area of ​​the non-Newtonian fluid in the action area of ​​the ultrasonic generator is the area where the specific point of laser irradiation is located.

[0242] When the ultrasonic wave emitted by the ultrasonic generator is a low-frequency ultrasonic wave (ultrasonic wave is 20 to 100 Hz), the timing curve mainly reflects the macroscopic rheological properties of the non-Newtonian fluid; when the ultrasonic wave emitted by the ultrasonic generator is a medium-frequency ultrasonic wave (ultrasonic wave is 100 to 1000 Hz), the timing curve mainly reflects the microstructural changes of the non-Newtonian fluid; when the ultrasonic wave emitted by the ultrasonic generator is a high-frequency ultrasonic wave (ultrasonic wave is greater than 1000 Hz), the timing curve mainly reflects the molecular-level structural characteristics of the non-Newtonian fluid.

[0243] In other embodiments, the external force application module 202 further includes a control circuit for controlling the vibration intensity, frequency, and duration of the vibrator and / or the ultrasonic generator, and for controlling the intensity and duration of the impact of the impactor. The processor can be connected to the control circuit to send a quantitative impact application command and a vibration application command to the control circuit.

[0244] In some embodiments, the external force application module 202 further includes a stirring device, and the stirring device is used to stir the non-Newtonian fluid.

[0245] In some embodiments, the data acquisition module 203 includes a high-speed data acquisition card, a filtering and amplifying circuit, and a temperature sensor. The sampling rate of the high-speed data acquisition card is not less than 1 kHz. The filtering and amplifying circuit is used to improve signal quality, remove environmental noise and system interference, and the temperature sensor is used to detect the temperature of the non-Newtonian fluid in real time.

[0246] In some embodiments, the terminal device 204 also includes a display screen, an input keyboard and a communication interface, wherein the display screen is used to display the measurement process and the performance parameters; the input keyboard is used to set the detection process; and the communication interface is used to support connection with the control system or data acquisition module 203.

[0247] The performance detection system for non-Newtonian fluids needs to be calibrated using standard non-Newtonian fluids. During the calibration process, the data needs to be detected in combination with current standard instruments to ensure the accuracy and reliability of the algorithm.

[0248] In some embodiments, the performance detection system of the non-Newtonian fluid can also be provided with an electrostatic adsorption self-cleaning window, a tungsten-polyethylene composite protective layer and a wide temperature range constant temperature system based on the Peltier effect, wherein the electrostatic adsorption self-cleaning window is used for dust removal, the tungsten-polyethylene composite protective layer is used for shielding radiation, and the wide temperature range constant temperature system based on the Peltier effect is used to control the temperature of the non-Newtonian fluid.

[0249] A fourth aspect of the embodiments of the present application provides a terminal device, the principle block diagram of the terminal device can be as follows: Figure 4 As shown. The terminal device includes a processor, a memory, a network interface, a display screen and a temperature sensor connected via a system bus. The processor is used to provide computing and control capabilities. The memory includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the terminal device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a performance detection method for a non-Newtonian fluid is implemented. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the temperature sensor is pre-set inside the terminal device to detect the operating temperature of the internal device.

[0250] Those skilled in the art will understand that Figure 4 The principle block diagram shown in the figure is only a block diagram of a partial structure related to the solution of the present invention, and does not constitute a limitation on the terminal device to which the solution of the present invention is applied. The specific terminal device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0251] In some embodiments, an embodiment of the present application provides a terminal device, which includes a processor and a memory, the memory being used to store a computer program, and the processor being used to call and run the computer program stored in the memory to execute the steps of the non-Newtonian fluid performance detection method provided in the first aspect of the above-mentioned embodiment of the present application.

[0252] In a fifth aspect of an embodiment of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium is used to store a computer program, and the computer program enables a computer to execute the steps of the non-Newtonian fluid performance detection method provided in the first aspect of the embodiment of the present application.

[0253] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the above-described method embodiments. Any reference to memory, storage, database, or other media used in the various embodiments provided herein may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct RAMbus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM).

[0254] The technical features of the above embodiments can be combined without changing the basic principles of this application. In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0255] The above embodiments merely illustrate several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent application. It should be noted that a person skilled in the art could make various modifications and improvements without departing from the spirit of the present application, all of which fall within the scope of protection of the present application. Therefore, the scope of patent protection for the present application shall be determined by the appended claims.

Claims

1. A method for detecting the performance of a non-Newtonian fluid, characterized in that: include: When the surface of the non-Newtonian fluid is in a stationary state, obtaining a reference height of the surface of the non-Newtonian fluid; When the surface of the non-Newtonian fluid is in a non-stationary state, acquiring a height dataset of the surface of the non-Newtonian fluid based on a time series; constructing a timing curve based on the reference altitude and the altitude data set; Performing feature extraction processing on the timing curve to obtain at least one feature parameter; Using a preset algorithm to detect and process the at least one characteristic parameter to obtain performance parameters of the non-Newtonian fluid; The step of detecting and processing the at least one characteristic parameter using a preset algorithm to obtain the performance parameter of the non-Newtonian fluid includes: detecting and processing the at least one characteristic parameter using a yield stress model of the preset algorithm to obtain a yield stress parameter of the performance parameter; The yield stress model of the preset algorithm is used to detect and process the at least one characteristic parameter, and the step of obtaining the yield stress parameter of the performance parameter includes: using the first hidden layer of the yield stress model to perform feature extraction processing on the at least one characteristic parameter to obtain a first eigenvector; using the first target activation function of the yield stress model to perform nonlinear transformation processing on the first eigenvector to obtain a second eigenvector; using the second hidden layer of the yield stress model to perform feature extraction processing on the second eigenvector to obtain a third eigenvector; using the first target activation function to perform nonlinear transformation processing on the third eigenvector to obtain a fourth eigenvector; using the third hidden layer of the yield stress model to perform feature extraction processing on the fourth eigenvector to obtain a fifth eigenvector; using the first target activation function to perform nonlinear transformation processing on the fifth eigenvector to obtain the yield stress parameter.

2. The method for detecting the performance of a non-Newtonian fluid according to claim 1, wherein: The step of constructing a timing curve based on the reference altitude and the altitude data set comprises: Calculating the difference between the reference height and each height data in the height data set to obtain target height change data corresponding to each height data; The time series curve is constructed based on all target height change data.

3. The method for detecting the performance of a non-Newtonian fluid according to claim 1, wherein: The step of detecting and processing the at least one characteristic parameter using a preset algorithm to obtain the performance parameter of the non-Newtonian fluid further includes: The plastic viscosity model of the preset algorithm is used to detect and process the at least one characteristic parameter to obtain the plastic viscosity parameter of the performance parameter.

4. The method for detecting the performance of a non-Newtonian fluid according to claim 1, wherein: The method for detecting the performance of a non-Newtonian fluid further includes a step of training the yield stress model, and the step of training the yield stress model includes: Acquire a historical data set and a first true label set, wherein each first true label in the first true label set has a one-to-one correspondence with one of the historical data in the historical data set; Using the original yield stress model to detect and process target historical data to obtain a first prediction result, the target historical data being any historical data in the historical data set; Constructing a first loss function based on the first prediction result and a first true label corresponding to the target historical data; The first model parameter of the original yield stress model is adjusted based on the first loss function and the first preset constraint item to form the yield stress model.

5. The method for detecting the performance of a non-Newtonian fluid according to claim 4, wherein: The yield stress model training step further includes: Using a target evaluation method to evaluate the yield stress model to obtain a first evaluation result; A second model parameter of the yield stress model is adjusted based on the first evaluation result.

6. A performance detection device for non-Newtonian fluids, characterized in that: include: A reference height acquisition module, configured to acquire a reference height of the surface of the non-Newtonian fluid when the surface of the non-Newtonian fluid is in a stationary state; a height dataset acquisition module, configured to acquire a height dataset of the surface of the non-Newtonian fluid based on a time sequence when the surface of the non-Newtonian fluid is in a non-stationary state; a timing curve construction module, configured to construct a timing curve based on the reference altitude and the altitude data set; A feature extraction module, configured to perform feature extraction processing on the timing curve to obtain at least one feature parameter; a performance parameter acquisition module, configured to detect and process the at least one characteristic parameter using a preset algorithm to obtain the performance parameter of the non-Newtonian fluid; The performance parameter acquisition module is further configured to detect and process the at least one characteristic parameter using the yield stress model of the preset algorithm to obtain the yield stress parameter of the performance parameter; The performance parameter acquisition module is further used to use the first hidden layer of the yield stress model to perform feature extraction processing on the at least one characteristic parameter to obtain a first eigenvector; use the first target activation function of the yield stress model to perform nonlinear transformation processing on the first eigenvector to obtain a second eigenvector; use the second hidden layer of the yield stress model to perform feature extraction processing on the second eigenvector to obtain a third eigenvector; use the first target activation function to perform nonlinear transformation processing on the third eigenvector to obtain a fourth eigenvector; use the third hidden layer of the yield stress model to perform feature extraction processing on the fourth eigenvector to obtain a fifth eigenvector; and use the first target activation function to perform nonlinear transformation processing on the fifth eigenvector to obtain the yield stress parameter.

7. A performance detection system for non-Newtonian fluids, characterized in that: The invention comprises a laser ranging module, an external force application module, a data acquisition module and a terminal device, wherein the terminal device comprises a processor and a memory, the processor is communicatively connected with the laser ranging module, the external force application module and the data acquisition module, the memory is used to store a computer program, and the processor is used to call and run the computer program stored in the memory to execute the steps of the performance detection method of the non-Newtonian fluid according to any one of claims 1 to 5.

8. A terminal device, characterized in that: include: A processor and a memory, the memory being used to store a computer program, the processor being used to call and run the computer program stored in the memory to execute the steps of the method for detecting the performance of a non-Newtonian fluid as described in any one of claims 1 to 5.

9. A computer-readable storage medium, characterized in that Used to store a computer program, wherein the computer program enables a computer to execute the steps of the method for detecting the performance of a non-Newtonian fluid according to any one of claims 1 to 5.

Images