An online detection method for temperature field and ring thickness distribution in a rotary kiln

Abstract

The present invention provides an online detection method for temperature field and ring thickness distribution in a rotary kiln, belonging to the technical field of cement production. The method comprises: collecting historical data and converting dimensions of measurable process variables in the industrial field, and constructing a measurable process variable condition set according to the historical data; establishing a spatial rectangular coordinate system and solid geometry by computational fluid dynamics simulation technology, and calculating the three-dimensional temperature field data in the kiln under the measurable process variable condition set according to the rotary kiln mechanism model; sampling the three-dimensional temperature field data under each condition along the X, Y, and Z axes and combining the measurable process variable data under the same condition, training a deep neural network to establish a kiln temperature field soft measurement model based on a deep neural network; and establishing a kiln ring thickness soft measurement model based on the heat transfer equation based on the online detected kiln temperature. The present invention can solve the technical problem that the temperature field and ring thickness distribution in the rotary kiln cannot be detected.

Description

Technical Field

[0001] The invention relates to the technical field of cement production, in particular to an online detection method for temperature field and ring thickness distribution in a rotary kiln. Background Art

[0002] The cement industry is an important basic industry for the development of the national economy and has played an important role in improving people's livelihood, economic construction and national defense security. Cement production is characterized by high pollution, high energy consumption and high emissions, but it is also an important means of treating urban garbage and hazardous waste. The cement production process can be summarized as "two grindings and one burning", among which clinker calcination is the most critical production link. As the core thermal equipment of the clinker calcination system, the operating status of the rotary kiln is of great significance to ensuring production safety, improving product quality and achieving energy conservation and emission reduction.

[0003] The temperature field and ring thickness in the kiln are important indicators of the operating status of the rotary kiln. If the temperature of the firing zone is too low, the quality of the clinker will be reduced, while if the temperature of the firing zone is too high, it is easy to cause serious ring formation in the kiln, thereby damaging the kiln lining and causing safety accidents. However, due to the harsh and relatively closed internal environment of the rotary kiln, the existing technology cannot detect the temperature field and ring thickness distribution in the kiln in real time, which makes the operating status of the rotary kiln very uncertain. Summary of the invention

[0004] The embodiment of the present invention provides an online detection method for the temperature field and ring thickness distribution in a rotary kiln, which can solve the technical problem that the temperature field and ring thickness distribution in a rotary kiln cannot be detected. The technical solution is as follows:

[0005] On the one hand, an embodiment of the present invention provides an online detection method for temperature field and ring thickness distribution in a rotary kiln, which is applied to electronic equipment and includes:

[0006] Collect historical data and convert dimensions of measurable process variables on industrial sites, and construct a set of measurable process variable conditions based on historical data;

[0007] The spatial rectangular coordinate system and solid geometry are established through computational fluid dynamics simulation technology, and the three-dimensional temperature field data in the kiln under the measurable process variable working condition set is calculated according to the rotary kiln mechanism model;

[0008] The three-dimensional temperature field data under each working condition is sampled along the X, Y, and Z axes and combined with the measurable process variable data under the same working condition, a deep neural network is trained to establish a soft measurement model of the kiln temperature field based on the deep neural network, and the online detection of the temperature distribution in the kiln is realized;

[0009] Based on the online detected kiln temperature, a soft measurement model of the kiln ring thickness is established through the heat transfer equation to realize the online detection of the kiln ring thickness distribution.

[0010] Further, the industrial on-site measurable process variables include: burner axial flow wind pressure, burner swirl wind pressure, burner coal flow, material flow and secondary air temperature;

[0011] The historical data collection and dimension conversion of the measurable process variables on the industrial site and the construction of the measurable process variable condition set according to the historical data include:

[0012] Collect historical data of five industrial field measurable process variables: burner axial air pressure, burner swirl air pressure, burner coal flow, material flow and secondary air temperature;

[0013] The burner axial flow wind pressure and burner swirl wind pressure are converted into axial flow wind velocity and swirl wind velocity through the pressure-wind speed conversion formula;

[0014] Based on the collected historical data, a set of measurable process variable conditions including axial flow velocity, swirl flow velocity, burner coal flow, material flow and secondary air temperature is constructed:

[0015]

[0016] Among them, the subscripts min, mean and max represent the minimum, mean and maximum values ​​respectively; X1, X2, X3, X4 and X5 represent the value sets of axial flow velocity, swirl flow velocity, burner coal flow rate, material flow rate and secondary air temperature respectively; Φ represents the set of measurable process variable operating conditions.

[0017] Furthermore, the pressure-wind speed conversion formula is expressed as:

[0018]

[0019] Among them, C p Indicates the flow coefficient of the throttle valve; A t It represents the flow area of ​​the throttle valve port; △p represents the pressure difference between the inlet and outlet of the throttle valve; S represents the flow area of ​​the burner air duct; v represents the wind speed in the burner air duct.

[0020] Furthermore, the establishment of a spatial rectangular coordinate system and solid geometry by computational fluid dynamics simulation technology includes:

[0021] A spatial rectangular coordinate system is established with the center of the secondary air inlet plane as the coordinate origin O, the horizontal direction as the X-axis, the vertical direction as the Y-axis, and the rotary kiln axis direction as the Z-axis;

[0022] According to the inner diameter of the rotary kiln c , material filling angle ω, material repose angle β r, kiln body inclination tanθ and burner structure parameters are used to create solid geometry in 1:1 scale.

[0023] Furthermore, the rotary kiln mechanism model satisfies the equations of the law of conservation of mass, the law of conservation of energy and the law of conservation of momentum;

[0024] The rotary kiln mechanism model includes: a Realizable k-ε turbulence model for controlling the turbulent flow process of the fluid in the kiln, a P1 radiation model for controlling the radiation heat transfer process in the kiln, a non-premixed combustion probability density function model for controlling the non-premixed combustion process of the fuel and oxidant in the kiln, and a discrete phase model for controlling the diffusion process of the coal powder particles.

[0025] Furthermore, the established soft-sensing model of kiln temperature field based on deep neural network is expressed as:

[0026]

[0027] Among them, x1, x2, x3, x4, x5, x6, x7 and x8 represent the normalized axial wind velocity, swirl wind velocity, coal flow rate, material flow rate, secondary air temperature, X-axis coordinate, Y-axis coordinate and Z-axis coordinate respectively; represents the output of the i-th neuron in the first layer, represents the output of the jth neuron in the second layer,…, represents the output of the kth neuron in the l-1th layer; σ1 represents the activation function of the first layer, σ2 represents the activation function of the second layer, …, σ l-1 represents the activation function of the l-1th layer; represents the connection weight between the first input of the first layer and the i-th neuron of the second layer, represents the connection weight between the second input of the first layer and the i-th neuron of the second layer,…, represents the connection weight between the j-th input of the l-1th layer and the k-th neuron of the lth layer; represents the connection bias of the i-th neuron in the first layer, represents the connection bias of the jth neuron in the second layer,…, represents the connection bias of the kth neuron in the l-1th layer; y represents the temperature at a certain position in the kiln; i, j...k represent the current neuron numbers in the 1st, 2nd...l-1th layers respectively; n1, n2 ……n l-1 Represent the total number of neurons in the 1st, 2nd, ..., l-1th layers respectively.

[0028] Furthermore, the established soft-sensing model of kiln ring thickness is expressed as:

[0029]

[0030] Where λ represents the thermal conductivity of air; P r represents the Prandtl number; D represents the diameter of the kiln shell; υ represents the air viscosity; n represents the kiln speed; V represents the ambient wind speed; g represents the acceleration of gravity; β represents the thermal expansion coefficient of air, which is 2 / (T m +T a );α represents the convection heat transfer coefficient of the kiln shell; represents the heat transfer of the kiln shell over the length dl; σ represents the Boltzmann constant; ε e Indicates the emissivity of the kiln shell; T a 、T m and T w Respectively represent the ambient temperature, kiln shell temperature and kiln temperature; m , b and λ c Respectively represent the thermal conductivity of the kiln shell, the thermal conductivity of the refractory bricks and the thermal conductivity of the ring; r m 、r b 、r c They respectively represent the outer surface radius of the kiln shell, the outer surface radius of the refractory brick, and the outer surface radius of the ring; d represents the thickness of the ring.

[0031] On the one hand, an electronic device is provided, comprising a processor and a memory, wherein the memory stores at least one instruction, and the at least one instruction is loaded and executed by the processor to implement the above-mentioned online detection method of the temperature field and ring thickness distribution in the rotary kiln.

[0032] On the one hand, a computer-readable storage medium is provided, wherein at least one instruction is stored in the storage medium, and the at least one instruction is loaded and executed by a processor to implement the above-mentioned online detection method of the temperature field and ring thickness distribution in the rotary kiln.

[0033] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0034] In the embodiment of the present invention, X, Y, and Z axis coordinates are introduced into the input of the soft measurement model of the temperature field in the kiln, so as to realize the online detection of the temperature at any position in the three-dimensional space through the measurable process variables, and the soft measurement calculation results of the temperature in the kiln are used to realize the online detection of the ring thickness in the kiln, so as to solve the technical problem that the temperature field and the ring thickness distribution in the rotary kiln cannot be detected, thereby more accurately evaluating the operating status of the rotary kiln. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0036] Figure 1 A schematic flow chart of an online detection method for temperature field and ring thickness distribution in a rotary kiln provided by an embodiment of the present invention;

[0037] Figure 2 A schematic diagram of the principle of an online detection method for temperature field and ring thickness distribution in a rotary kiln provided by an embodiment of the present invention;

[0038] Figure 3 A three-dimensional geometrical schematic diagram of a rotary kiln provided in an embodiment of the present invention;

[0039] Figure 4 A schematic diagram of various media in a rotary kiln shaft cross section provided by an embodiment of the present invention;

[0040] Figure 5 It is a structural schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0041] In order to make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0042] like Figure 1 and Figure 2 As shown, an embodiment of the present invention provides an online detection method for temperature field and ring thickness distribution in a rotary kiln, which can be implemented by an electronic device, and includes:

[0043] S101, collecting historical data and converting dimensions of measurable process variables on industrial sites, and constructing a measurable process variable condition set based on the historical data;

[0044] In this embodiment, the industrial on-site measurable process variables include: burner axial flow wind pressure, burner swirl wind pressure, burner coal flow, material flow and secondary air temperature;

[0045] In this embodiment, the historical data collection and dimension conversion of the measurable process variables of the industrial site and the construction of the measurable process variable condition set according to the historical data may specifically include the following steps:

[0046] A1, historical data collection of five industrial field measurable process variables including burner axial air pressure, burner swirl air pressure, burner coal flow, material flow and secondary air temperature;

[0047] A2, convert the burner axial flow wind pressure and the burner swirl wind pressure into the axial flow wind velocity and the swirl wind velocity through the pressure-wind speed conversion formula; wherein the pressure-wind speed conversion formula is expressed as:

[0048]

[0049] Among them, C p Indicates the flow coefficient of the throttle valve; A t represents the flow area of ​​the throttle valve port; △p represents the pressure difference between the inlet and outlet of the throttle valve; S represents the flow area of ​​the burner air duct; v represents the wind speed in the burner air duct;

[0050] A3, based on the collected historical data, construct a set of measurable process variable conditions including axial flow velocity, swirl flow velocity, burner coal flow, material flow and secondary air temperature:

[0051]

[0052] Among them, the subscripts min, mean and max represent the minimum, mean and maximum values ​​respectively; X1, X2, X3, X4 and X5 represent the value sets of axial flow velocity, swirl flow velocity, burner coal flow rate, material flow rate and secondary air temperature respectively; Φ represents the set of measurable process variable operating conditions.

[0053] S102, establishing a spatial rectangular coordinate system and solid geometry by computational fluid dynamics simulation technology, and calculating three-dimensional temperature field data in the kiln under a set of measurable process variable conditions according to a rotary kiln mechanism model, wherein the three-dimensional temperature field data includes: X, Y, Z axis coordinates and temperature values ​​of corresponding coordinates (specifically: temperature in the kiln);

[0054] In this embodiment, Figure 3 As shown, the spatial rectangular coordinate system and the method for establishing the solid geometry include: establishing a spatial rectangular coordinate system with the center of the secondary air inlet plane as the coordinate origin O, the horizontal direction as the X-axis, the vertical direction as the Y-axis, and the axis direction of the rotary kiln as the Z-axis; according to the inner diameter r of the rotary kiln c , material filling angle ω, material repose angle β, kiln body inclination tanα, burner structure and other parameters are used to create solid geometry in a 1:1 scale.

[0055] In this embodiment, the rotary kiln mechanism model includes: a Realizable k-ε turbulence model (i.e., a realizable k-ε turbulence model) for controlling the turbulent flow process of the fluid in the kiln, a P1 radiation model for controlling the radiation heat transfer process in the kiln, a non-premixed combustion probability density function (Probability Density Function, PDF) model for controlling the non-premixed combustion process of the fuel and oxidant in the kiln, and a discrete phase (Discrete Phase Model, DPM) model for controlling the diffusion process of the coal powder particles. In addition, the rotary kiln mechanism model should also satisfy the mass conservation law, energy conservation law, and momentum conservation law equations.

[0056] S103, sampling the three-dimensional temperature field data under each working condition along the X, Y, and Z axes and combining the measurable process variable data under the same working condition, training a deep neural network to establish a kiln temperature field soft measurement model based on the deep neural network, and realizing online detection of the temperature distribution in the kiln;

[0057] In this embodiment, the input data is normalized and preprocessed by formula (3):

[0058]

[0059] Among them, x i-input Represents input data (including: axial wind velocity, swirl wind velocity, coal flow, material flow, secondary air temperature, X-axis coordinate, Y-axis coordinate and Z-axis coordinate); i-min and x i-max Represents the input data x i-input The maximum and minimum values ​​of x i Represents normalized data.

[0060] In this embodiment, the forward propagation process of the deep neural network is realized by formula (4):

[0061]

[0062] in, represents the output of the kth neuron in the lth layer of the deep neural network; σ l represents the activation function of the lth layer; m represents the number of neurons in the l-1th layer; represents the connection weight between the jth neuron in the l-1th layer and the kth neuron in the lth layer; represents the output of the jth neuron in the l-1th layer; Represents the connection bias of the k-th neuron in the l-th layer;

[0063] In this embodiment, the loss function is calculated by formula (5):

[0064]

[0065] Where θ represents the updated weight parameter; J(θ) represents the loss function with parameter θ; h represents the number of samples in each batch; y i Represents the output prediction value; o i Indicates the output target value.

[0066] In this embodiment, the weight parameters of the neural network are updated by formula (6):

[0067]

[0068] Where s represents the number of steps to update; θ represents the updated weight parameter; J(θ) represents the loss function with parameter θ; g s Denotes the loss function J s (θ s-1 ) is the gradient obtained by taking the derivative of θ; β1 represents the first-order moment attenuation coefficient; β2 represents the second-order moment attenuation coefficient; m s represents the gradient g s The first moment of v s represents the gradient g s The second moment of represents the bias-corrected value of the first-order moment; represents the bias correction value of the second-order moment; α represents the learning rate; ε0 is the stability coefficient;

[0069] In this embodiment, whether the training process is stopped is determined by formula (7):

[0070] |J(θ s+1 )-J(θ s )|<ξ (7)

[0071] Here, ξ is a user-defined value.

[0072] The soft-sensing model of kiln temperature field based on deep neural network after training can be expressed as:

[0073]

[0074] Among them, x1, x2, x3, x4, x5, x6, x7 and x8 represent the normalized axial wind velocity, swirl wind velocity, coal flow rate, material flow rate, secondary air temperature, X-axis coordinate, Y-axis coordinate and Z-axis coordinate respectively; represents the output of the i-th neuron in the first layer, represents the output of the jth neuron in the second layer,…, represents the output of the kth neuron in the l-1th layer, and so on; σ1 represents the activation function of the first layer, σ2 represents the activation function of the second layer, …, σ l-1 represents the activation function of the l-1th layer, and so on; represents the connection weight between the first input of the first layer and the i-th neuron of the second layer, represents the connection weight between the second input of the first layer and the i-th neuron of the second layer,…, represents the connection weight between the jth input of the l-1th layer and the kth neuron of the lth layer, and so on; represents the connection bias of the i-th neuron in the first layer, represents the connection bias of the jth neuron in the second layer,…, represents the connection bias of the kth neuron in the l-1th layer, and so on; y represents the temperature at a certain position in the kiln; i, j...k represent the current neuron numbers in the 1st, 2nd...l-1th layers respectively; n1, n2 ……n l-1 Represent the total number of neurons in the 1st, 2nd, ..., l-1th layers respectively.

[0075] S104, based on the online detected kiln temperature and kiln shell temperature, ambient temperature and ambient wind speed, a soft measurement model of the kiln ring thickness is established through the heat transfer equation to achieve online detection of the kiln ring thickness distribution. Specifically, the following steps may be included:

[0076] In this embodiment, the convective heat transfer coefficient of the kiln shell is calculated by equations (9) to (12):

[0077]

[0078] Where D is the diameter of the kiln shell; υ a Re represents air viscosity; n represents kiln speed (unit: rpm); Re w represents the rotational Reynolds number; V represents the ambient wind speed; Re represents the cross-flow Reynolds number; T m Indicates the outer surface temperature of the kiln shell (abbreviated as: kiln shell temperature); T a Indicates the ambient temperature; β a Represents the thermal expansion coefficient of air, which is 2 / (T m +T a ), g represents the acceleration due to gravity, Gr represents the Grashof number, λ represents the thermal conductivity of air, P r represents the Prandtl number; α c It represents the convection heat transfer coefficient of the kiln shell;

[0079] In this embodiment, the heat dissipation of the kiln shell is calculated by equations (13) to (15):

[0080]

[0081] in, and They represent the radiation heat transfer, convection heat transfer and total heat transfer of the kiln shell over the length dl respectively; r mIndicates the radius of the kiln shell; α c represents the convective heat transfer coefficient of the kiln shell; σ represents the Boltzmann constant; ε e Indicates the emissivity of the kiln shell; T a and T m Represent the ambient temperature and the outer surface temperature of the kiln shell respectively.

[0082] In this embodiment, the heat transfer relationship through different medium layers of the kiln wall can be expressed as:

[0083]

[0084] in, represents the heat transfer of the kiln shell over the length dl; T m 、T b 、T c and T w Respectively represent the outer surface temperature of the kiln shell, the outer surface temperature of the refractory bricks, the outer surface temperature of the ring and the inner surface temperature of the ring (kiln temperature); m , b and λ c Respectively represent the thermal conductivity of the kiln shell, the thermal conductivity of the refractory bricks and the thermal conductivity of the ring; r m 、r b 、r c and r w They represent the outer radius of the kiln shell, the outer radius of the refractory brick, the outer radius of the ring and the inner radius of the ring respectively. Figure 4 As shown;

[0085] In this embodiment, the simplified soft-sensing model of the ring thickness in the kiln can be expressed as:

[0086]

[0087] Where λ represents the thermal conductivity of air; P r represents the Prandtl number; D represents the diameter of the kiln shell; υ a represents air viscosity; n represents kiln speed (unit: rpm); V represents ambient wind speed; g represents gravitational acceleration; β a Represents the thermal expansion coefficient of air, which is 2 / (T m +T a ); α c It represents the convection heat transfer coefficient of the kiln shell; represents the heat transfer of the kiln shell over the length dl; σ represents the Boltzmann constant; ε e Indicates the emissivity of the kiln shell; T a 、T m and T w Respectively represent the ambient temperature, kiln shell temperature and the inner surface temperature of the ring (i.e.: kiln temperature); λ m ,b and λ c Respectively represent the thermal conductivity of the kiln shell, the thermal conductivity of the refractory bricks and the thermal conductivity of the ring; r m 、r b 、r c They respectively represent the outer surface radius of the kiln shell, the outer surface radius of the refractory brick, and the outer surface radius of the ring; d represents the thickness of the ring.

[0088] In summary, the beneficial effects brought about by the technical solution provided by the present invention include at least:

[0089] 1. The X, Y, and Z axis coordinates are introduced into the soft measurement model input of the temperature field in the kiln, which realizes the online detection of the temperature at any position in the three-dimensional space through the measurable process variables, and can visualize the temperature field distribution, which has good intuitiveness and interactivity;

[0090] 2. The soft measurement model of the thickness of the kiln ring can realize the online detection of the thickness of the kiln ring by using the measurable process variables (including: kiln shell temperature, ambient temperature and ambient wind speed) and the soft measurement calculation results of the kiln temperature output by the soft measurement model of the kiln temperature field, and can visualize the distribution of the ring thickness, which has good accuracy and practicality;

[0091] 3. The online detection method for the temperature field and ring thickness distribution in the rotary kiln provided in the embodiment of the present invention is universal and suitable for the problem that variables cannot be detected in the industrial process. Even when the acquired sensor data is very limited, the present invention can still be executed normally. The method for solving the problem that the temperature field and ring distribution in the rotary kiln cannot be detected, the detection result reduces the uncertainty of the industrial process, and can be used to judge the operating status of the rotary kiln, optimize process parameters and assist manual decision-making; the present invention can be applicable to a wider range of industrial scenarios, making the industrial process more precise and intelligent.

[0092] Figure 5 It is a structural schematic diagram of an electronic device 600 provided in an embodiment of the present invention. The electronic device 600 may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 601 and one or more memories 602, wherein at least one instruction is stored in the memory 602, and the at least one instruction is loaded and executed by the processor 601 to realize the above-mentioned online detection method of temperature field and ring thickness distribution in the rotary kiln.

[0093] In an exemplary embodiment, a computer-readable storage medium is also provided, such as a memory including instructions, which can be executed by a processor in a terminal to complete the above-mentioned online detection method of temperature field and ring thickness distribution in a rotary kiln. For example, the computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0094] A person skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware or by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.

[0095] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An online detection method for temperature field and ring thickness distribution in a rotary kiln, characterized in that: include: Collect historical data and convert dimensions of measurable process variables on industrial sites, and construct a set of measurable process variable conditions based on historical data; The spatial rectangular coordinate system and solid geometry are established through computational fluid dynamics simulation technology, and the three-dimensional temperature field data in the kiln under the measurable process variable working condition set is calculated according to the rotary kiln mechanism model; The three-dimensional temperature field data under each working condition is sampled along the X, Y, and Z axes and combined with the measurable process variable data under the same working condition, a deep neural network is trained to establish a soft measurement model of the kiln temperature field based on the deep neural network, and the online detection of the temperature distribution in the kiln is realized; Based on the online detected kiln temperature, a soft measurement model of the kiln ring thickness is established through the heat transfer equation to achieve online detection of the kiln ring thickness distribution; Among them, the established soft-sensing model of kiln temperature field based on deep neural network is expressed as: Among them, x1, x2, x3, x4, x5, x6, x7 and x8 represent the normalized axial wind velocity, swirl wind velocity, coal flow rate, material flow rate, secondary air temperature, X-axis coordinate, Y-axis coordinate and Z-axis coordinate respectively; represents the output of the i-th neuron in the first layer, represents the output of the jth neuron in the second layer,…, represents the output of the kth neuron in the l-1th layer; σ1 represents the activation function of the first layer, σ2 represents the activation function of the second layer, …, σ l-1 represents the activation function of the l-1th layer; represents the connection weight between the first input of the first layer and the i-th neuron of the second layer, represents the connection weight between the second input of the first layer and the i-th neuron of the second layer,…, represents the connection weight between the j-th input of the l-1th layer and the k-th neuron of the lth layer; represents the connection bias of the ith neuron in the first layer, represents the connection bias of the jth neuron in the second layer,…, represents the connection bias of the kth neuron in the l-1th layer; y represents the temperature at a certain position in the kiln; i, j...k represent the current neuron numbers in the 1st, 2nd...l-1th layers respectively; n1, n2...n l-1 Respectively represent the total number of neurons in the 1st, 2nd, ..., l-1th layers; Among them, the established soft measurement model of kiln ring thickness is expressed as: Where λ represents the thermal conductivity of air; P r represents the Prandtl number; D represents the diameter of the kiln shell; υ represents the air viscosity; n represents the kiln speed; V represents the ambient wind speed; g represents the acceleration of gravity; β represents the thermal expansion coefficient of air, which is 2 / (T m +T a );α represents the convection heat transfer coefficient of the kiln shell; represents the heat transfer of the kiln shell over the length dl; σ represents the Boltzmann constant; ε e Indicates the emissivity of the kiln shell; T a 、T m and T w Respectively represent the ambient temperature, kiln shell temperature and kiln temperature; m , b and λ c Respectively represent the thermal conductivity of the kiln shell, the thermal conductivity of the refractory bricks and the thermal conductivity of the ring; r m 、r b 、r c They respectively represent the outer surface radius of the kiln shell, the outer surface radius of the refractory brick, and the outer surface radius of the ring; d represents the thickness of the ring.

2. The method for online detection of temperature field and ring thickness distribution in a rotary kiln according to claim 1, characterized in that: The industrial on-site measurable process variables include: burner axial flow wind pressure, burner swirl wind pressure, burner coal flow, material flow and secondary air temperature; The historical data collection and dimension conversion of the measurable process variables on the industrial site and the construction of the measurable process variable condition set according to the historical data include: Collect historical data of five industrial field measurable process variables: burner axial air pressure, burner swirl air pressure, burner coal flow, material flow and secondary air temperature; The burner axial flow wind pressure and burner swirl wind pressure are converted into axial flow wind velocity and swirl wind velocity through the pressure-wind speed conversion formula; Based on the collected historical data, a set of measurable process variable conditions including axial flow velocity, swirl flow velocity, burner coal flow, material flow and secondary air temperature is constructed: Among them, the subscripts min, mean and max represent the minimum, mean and maximum values ​​respectively; X1, X2, X3, X4 and X5 represent the value sets of axial flow velocity, swirl flow velocity, burner coal flow rate, material flow rate and secondary air temperature respectively; Φ represents the set of measurable process variable operating conditions.

3. The method for online detection of temperature field and ring thickness distribution in a rotary kiln according to claim 2, characterized in that: The pressure-wind speed conversion formula is expressed as: Among them, C p Indicates the flow coefficient of the throttle valve; A t It represents the flow area of ​​the throttle valve port; △p represents the pressure difference between the inlet and outlet of the throttle valve; S represents the flow area of ​​the burner air duct; v represents the wind speed in the burner air duct.

4. The method for online detection of temperature field and ring thickness distribution in a rotary kiln according to claim 1, characterized in that: The establishment of a spatial rectangular coordinate system and solid geometry by computational fluid dynamics simulation technology includes: A spatial rectangular coordinate system is established with the center of the secondary air inlet plane as the coordinate origin O, the horizontal direction as the X-axis, the vertical direction as the Y-axis, and the rotary kiln axis direction as the Z-axis; According to the inner diameter of the rotary kiln c , material filling angle ω, material repose angle β r , kiln body inclination tanθ and burner structure parameters are used to create solid geometry in 1:1 scale.

5. The method for online detection of temperature field and ring thickness distribution in a rotary kiln according to claim 1, characterized in that: The rotary kiln mechanism model satisfies the equations of the law of conservation of mass, the law of conservation of energy and the law of conservation of momentum; The rotary kiln mechanism model includes: a Realizable k-ε turbulence model for controlling the turbulent flow process of the fluid in the kiln, a P1 radiation model for controlling the radiation heat transfer process in the kiln, a non-premixed combustion probability density function model for controlling the non-premixed combustion process of the fuel and oxidant in the kiln, and a discrete phase model for controlling the diffusion process of the coal powder particles.

Images