Adaptive temperature control electric heat tracing band and control method

Abstract

The application discloses a self-adaptive temperature control electric tracing band, which comprises an electric tracing band body, a built-in distributed optical fiber and a smart slave station unit based on power line carrier communication / supply integration; a control center, which comprises an edge gateway and a self-adaptive temperature control algorithm engine; the edge gateway is installed in a field control box and is responsible for driving an optical fiber sensing demodulator to acquire a full-line temperature field and vibration signals in real time; and the self-adaptive temperature control algorithm engine runs in the edge gateway and performs data fusion, state analysis, heat demand prediction and power decision. The self-adaptive temperature control electric tracing band can continuously track real changes of pipeline characteristics, thereby being capable of coping with various uncertainties and time-variations in a real engineering environment.

Description

Technical Field

[0001] This invention belongs to the field of electric heating tape technology, specifically, it relates to an adaptive temperature-controlled electric heating tape and its control method. Background Technology

[0002] Current electric heating tape technology is mainly divided into constant power type and self-regulating (self-limiting) type. Although they have been widely used, the following key technical bottlenecks still exist in achieving accurate, efficient, reliable temperature maintenance that can adapt to complex working conditions: Slow response and insufficient temperature control accuracy: Self-regulating electric heating tapes rely on the inherent physical properties of PTC (positive temperature coefficient) materials for "self-adjustment." This adjustment is passive and delayed, and its response speed is limited by the thermal inertia of the material. It cannot actively predict temperature change trends, resulting in significant temperature fluctuations (such as ±5℃ or even larger) in the heated pipes or equipment. This makes it difficult to meet the extremely high temperature stability requirements of scenarios such as precision chemical reactions and critical pipelines of high-speed railways in cold regions.

[0003] Poor temperature uniformity in long-distance and complex pipelines: For long-distance pipelines (especially series-connected constant power electric heating cables that can reach the kilometer level) or complex pipeline networks with numerous valves, flanges, supports, and other "heat sinks," traditional technologies struggle to achieve uniform heating throughout the entire line. Heat conduction along the pipeline axis exhibits a gradient, and localized heat loss is significant at the heat sink locations, easily forming "cold spots" that pose a risk of freezing or media solidification. Overcompensating for "cold spots," on the other hand, can lead to overheating in other areas, impacting safety and energy consumption.

[0004] Lack of Adaptive and Predictive Capabilities: Existing technologies lack the ability to perceive and adapt to environmental conditions (such as sudden changes in wind speed and ambient temperature), changes in medium state (such as changes in flow rate), and the aging of the system itself (such as performance degradation of the insulation layer and performance drift of PTC materials). The control logic is static or based on feedback at the current moment, and cannot make forward-looking adjustments. When there are large temperature differences between day and night, sudden weather changes, or changes in process conditions, the system becomes unstable.

[0005] Gap in fault diagnosis and performance degradation early warning: Faults in electric heat tracing systems (such as reduced insulation, local damage, and water ingress into joints) are usually only discovered after the heat tracing fails. There is a lack of online, non-invasive early diagnosis and early warning methods, resulting in passive and costly maintenance. Summary of the Invention

[0006] To address the shortcomings of existing technologies, the present invention aims to provide an adaptive temperature-regulating electric heating tape.

[0007] To achieve the aforementioned objectives, the technical solution adopted by this invention includes: An adaptive temperature-regulating electric heating tape, comprising: The electric heating cable itself has a built-in distributed optical fiber (DTS / DAS) and an intelligent slave unit based on power line carrier (PLC) communication / power supply integration (used to measure micro-area impedance R_i(t)). The sensing unit includes a distributed temperature sensor, a distributed acoustic / vibration sensor, and a fiber optic sensor demodulator. The single-mode fiber of the fiber optic sensor demodulator is embedded between the insulation layer and the shielding layer along the axis of the electric heating cable body, serving as both a signal transmission channel and a sensing medium for the distributed temperature sensor and the distributed acoustic / vibration sensor. Fiber optic connectors are provided at the power supply cold end and the remote terminal of the electric heating cable body. These connectors fusion-splice the single-mode fiber to an external standard single-mode fiber patch cord (LC / APC connector) for protection and achieve the same level of protection and sealing as the heating cable body, ensuring that the sensing signal is reliably led out of the fiber optic sensor demodulator. The intelligent slave unit integrates a passive intelligent slave unit during the manufacturing process of the electric heating cable at regular intervals (e.g., 30 meters). This module abandons the traditional wireless and independent power supply design, and innovatively adopts integrated power line carrier (PLC) communication and power supply technology. The module is directly connected to the power supply wire of the heating cable through a high-frequency coupling circuit, extracts the high-frequency carrier signal from the wire, and supplies power to its own circuit after rectification and voltage regulation; at the same time, this circuit is responsible for demodulating the commands issued by the master station and modulating and uploading data such as the equivalent impedance R_i(t) collected in the local area. The control center includes an edge gateway and an adaptive temperature control algorithm engine. The edge gateway, installed in the field control box, is responsible for driving the fiber optic sensor demodulator to acquire the temperature field and vibration signals of the entire line in real time; at the same time, it communicates with each intelligent slave unit to collect local electrical parameters (such as micro-area impedance); it is equipped with the following key interfaces: Fiber optic sensor demodulator interface: drives the demodulator to acquire the temperature field T(z,t) and vibration characteristics V(z,t) of the entire line in real time; Power line carrier master station modem: broadcasts control commands to all intelligent slave units via the heating cable itself, and receives data uploaded by each slave unit; Digital output module: used to control the relays of the zone power regulation unit. The adaptive temperature control algorithm engine runs within the edge gateway and incorporates multiple algorithm models proposed in this invention to perform data fusion, status analysis, heat demand prediction, and power decision-making. The zoned power regulation unit electrically divides the entire electric heating cable into several independent heating sections (corresponding to the location of the intelligent slave unit). Each heating section is connected to the main circuit of the control center through a solid-state relay (SSR) or intelligent power switch, and its on / off state and duty cycle receive independent commands from the control center. Electrical topology: adopts a topology of "AC total input → multiple independent branch outputs". After the main air switch, N independent circuits (N is the number of zones) are connected in parallel. Actuation device: each independent circuit consists of a solid-state relay (SSR), a branch protection circuit breaker, and a corresponding heating section of the heating cable connected in series. Control connection: the control terminal (low-voltage DC side) of each SSR is connected to the digital output module of the control center to receive the on / off (PWM) commands issued by it.

[0008] An adaptive temperature-regulating electric heating tape control method, employing an adaptive temperature-regulating electric heating tape system, includes the following steps: A1. Full-dimensional state perception and data fusion: The temperature distribution, vibration characteristics and equivalent impedance of the corresponding section are fused to construct a digital twin of the pipeline thermal state. Specifically, the edge gateway acquires the temperature distribution T(z,t) and vibration characteristics V(z,t) of the entire pipeline (1-meter spatial resolution) once per second using a fiber optic sensor demodulator. Simultaneously, the power line carrier (PLC) channel collects the segment equivalent impedance R_i(t) reported by each slave station. The adaptive temperature control algorithm engine integrates these three types of data with a known pipeline structure diagram (marked with valve and flange locations), a preset target temperature T_target, and the real-time ambient temperature T_env(t) acquired through the gateway to construct a digital twin of the pipeline's thermal state.

[0009] A2. Dynamic heat demand modeling and prediction: Run a dynamic heat flow balance and prediction model, and based on the current temperature, ambient temperature, and thermal resistance of the insulation layer, introduce a temperature change trend factor and a local disturbance factor to predict future heat demand. Specifically, real-time heat loss calculation: For each pipe unit (e.g., every 1 meter), the current heat dissipation rate is calculated based on its current temperature T(z,t), ambient temperature T_env(t), and dynamically updated total equivalent thermal resistance R_total(z,t). Q_loss(z,t) = (T(z,t) - T_env(t)) / R_total(z,t), where R_total(z,0) = R_total_init(z), whose initial value is calculated from the design drawing parameters. R_total(z,0) represents the total equivalent thermal resistance of the pipe at axial position z and time t=0 (initial moment), and R_total_init(z) represents the initial design calculation value of the total equivalent thermal resistance of the pipe at position z. It is a pre-calculated static array or function that is only related to position z.

[0010] Heat demand forecasting: Introducing a temperature change trend factor α(z,t)=[T(z,t)-T(z,t-Δt)] / Δt, where: T(z,t): the temperature value measured at position z at the end of the current control cycle t, T(z,t-Δt): the temperature value measured at position z at the end of the previous control cycle (t-Δt), Δt: the fixed control cycle of the system, i.e., the time interval between two consecutive executions of the control algorithm, [T(z,t)-T(z,t-Δt)]: the temperature change within one control cycle, and a local disturbance factor β(z,t) based on vibration analysis.

[0011] The predicted future heat demand is: the predicted linear power heat demand at position z and future time t+τ is: Q_demand_pred(z,t+τ)=Q_loss(z,t)*[1+γ1*α(z,t)+γ2*β(z,t)], where Q_loss(z,t) is the current heat dissipation rate, γ1 and γ2 are adaptive weight coefficients obtained through machine learning algorithm, and τ is the prediction step size.

[0012] For example, at a valve location, a sudden increase in the energy of a specific frequency band of V(z,t) may indicate a change in the medium flow rate, resulting in a corresponding increase in the β value and predicting an increase in heat demand at that point. The generation of β(z,t) involves first performing a short-time Fourier transform (STFT) on the vibration signal V(z,t) to obtain the time spectrum, and then calculating the signal energy E_band(z,t) of a specific frequency band (such as the band associated with fluid vortex shedding or valve throttling noise). β(z,t) can be obtained by normalizing the rate of change of this energy relative to the historical baseline energy, for example: β(z,t) = [E_band(z,t) - E_band_avg(z)] / E_band_avg(z), where E_band_avg(z) is the historical average energy at that point under steady state.

[0013] A3. Zonal adaptive power decision: Based on Q_demand_pred, calculate the required ideal heating power P_i_desired for each independent heating zone i. Specifically, the adaptive temperature control algorithm engine predicts the heat demand power Q_demand_pred per unit length and calculates the required ideal heating power P_i_desired for each independent heating segment i. Traditional PID control is replaced here by a collaborative algorithm of model predictive control (MPC) and fuzzy rules. With the optimization objective of minimizing the temperature deviation between the entire section and the target temperature, and minimizing the total energy consumption, a constrained optimization problem is solved in each control cycle, and the optimal control sequence of power switching of each section in the next few cycles is directly output (the first sequence is executed). To address model uncertainties (such as changes in R_ins caused by moisture in the insulation layer), a set of fuzzy rules is designed. For example, if T(z) in a certain section is consistently lower than T_target AND the power command for that section has been consistently high AND its R_i(t) does not show any abnormalities, then the "suspected insulation failure" flag is triggered, and the calculation weight of P_i_desired for that section is temporarily and slightly increased; A4. Control commands are issued to the power regulation units of each zone for execution. At the same time, the system enters the feedback learning phase. If a systematic shift in the curve is detected, it is determined that the PTC material in that section has aged or undergone local microstructural changes. The power-temperature transfer function of that section is then corrected online during power decision-making.

[0014] Specifically, the feedback learning phase includes: Model parameter self-calibration: The actual measured temperature T(z,t+1) at the next time step is compared with the MPC model prediction T_pred(z,t+1) for the temperature at position z at time t+1 (i.e., at the end of the next control cycle). Using backpropagation of the error, the total equivalent thermal resistance R_total(z,t+1) of each pipe element at position z used for control calculations at time t+1 and thereafter is updated online using algorithms such as recursive least squares (RLS). The updated parameters are immediately used in step A2 of the next control cycle, forming a closed-loop learning process.

[0015] Online calibration of PTC characteristics: Analyze the relationship curve between R_i(t) and average temperature T_avg_i(t) for each segment and compare it with the theoretical RT curve of the PTC material. If a systematic shift in the curve is found, it is determined that the PTC material in that segment has aging or local microstructural changes, and the power-temperature transfer function of that segment is corrected online during power decision-making.

[0016] A computer medium storing a computer program that, when executed, implements the steps of an adaptive adjustment method for controlling an electric heating cable.

[0017] Compared with the prior art, the advantages of the present invention include: (1) The present invention provides an adaptive temperature control electric heating tape and control method, which significantly improves the temperature control accuracy and uniformity: through 1-meter resolution full-area temperature sensing and MPC prediction optimization, the temperature fluctuation of long pipelines is reduced from more than ±5℃ in traditional technology to within ±1℃, effectively eliminating "cold spots".

[0018] (2) The present invention provides an adaptive temperature-controlled electric heating tape and control method, which significantly improves the system energy efficiency: based on the feedforward prediction of the dynamic thermal resistance model and MPC global optimization, the heating power and real-time heat demand are accurately matched. Compared with the traditional self-limiting heating tape, the energy saving is expected to be 25%-40%; (3) The adaptive temperature control electric heating tape and control method provided by the present invention have strong adaptive and anti-disturbance capabilities: the process disturbance (such as sudden flow change) is sensed by vibration signal, and the environmental changes (such as strong wind and rain) and system gradual changes (such as insulation layer getting damp) are adapted by fuzzy rules and online parameter correction (R_total update), and the system is extremely robust. (4) The present invention provides an adaptive temperature control electric heating tape and control method to achieve predictive health management: by online calibration of PTC characteristics, early warning of material aging can be given; by monitoring the abnormal increase of R_total(z,t), the insulation layer damage or water ingress fault can be located, changing "passive maintenance" to "active early warning", greatly improving system reliability and reducing maintenance costs; (5) The adaptive temperature control electric heating tape and control method provided by the present invention have strong engineering feasibility: the power line carrier communication, industrial standard optical fiber sensing and distributed SSR control used are all mature technologies. Through the unique architecture and algorithm of the present invention, they are innovatively integrated to form a complete solution with outstanding substantial progress. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart of an adaptive temperature-regulating electric heating tape and control method according to the present invention; Figure 2 This is a schematic diagram of the closed-loop logic of prediction-measurement-correction for an adaptive temperature-controlled electric heating tape and control method in this invention. Detailed Implementation

[0021] In view of the shortcomings of the prior art, the inventors of this invention, through long-term research and extensive practice, have proposed the technical solution of this invention. The technical solution, its implementation process, and principles will be further explained below with reference to the accompanying drawings and specific implementation examples in the embodiments of this application.

[0022] It should be noted that the embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention. The described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, the present invention covers any substitutions, modifications, equivalent methods and solutions made on the spirit, principles and scope of the present invention as defined by the claims. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] In the description of this application, the terms "first," "second," "third," and similar words do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Similarly, the terms "a" or "one," and similar words, do not indicate a quantity limitation, but rather indicate the presence of at least one. The terms "comprising" or "including," and similar words, mean that the elements or objects preceding "comprising" or "including" encompass the elements or objects listed following "comprising" or "including," and their equivalents, but do not exclude other elements or objects. The terms "connected" or "linked," and similar words, are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect.

[0024] In the description of this application, the terms "center," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used solely for the convenience of describing this application and for simplification, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application. Furthermore, when using positional terms such as "both sides," "outer side," and "upper and lower," it should be understood that they are used only for ease of understanding and description, taking into account that the structure may be oriented to other positions.

[0025] In the description of this application, unless otherwise expressly specified and limited, the technical or scientific terms used shall have the ordinary meaning understood by a person with ordinary skills in the art to which this application pertains. Terms such as “installation,” “connection,” and “joining” shall be interpreted broadly, for example, as fixed connection, detachable connection, mating connection, or integral connection. For a person skilled in the art, the specific meaning of the above terms in this application can be understood according to the specific circumstances.

[0026] The present invention aims to introduce and explain the structural composition of an adaptive temperature-regulating electric heating tape and control method, as well as the coordination relationship between the various components. Unless otherwise specified, the dimensions, materials, and manufacturing processes of the components suitable for the adaptive temperature-regulating electric heating tape and control method in the present invention can be selected according to specific circumstances, and no special limitations or explanations are given here.

[0027] Furthermore, to provide the public with a better understanding of the present invention, certain specific details are described in detail in the following description of the invention. However, those skilled in the art will fully understand the invention even without these detailed descriptions.

[0028] Example 1

[0029] Please see Figure 1 and Figure 2 An adaptive temperature-regulating electric heating tape, comprising: The electric heating cable itself has a built-in distributed optical fiber (DTS / DAS) and an intelligent slave unit based on power line carrier (PLC) communication / power supply integration (used to measure micro-area impedance R_i(t)). The sensing unit includes a distributed temperature sensor, a distributed acoustic / vibration sensor, and a fiber optic sensor demodulator. The single-mode fiber of the fiber optic sensor demodulator is embedded between the insulation layer and the shielding layer along the axis of the electric heating cable body, serving as both a signal transmission channel and a sensing medium for the distributed temperature sensor and the distributed acoustic / vibration sensor. Fiber optic connectors are provided at the power supply cold end and the remote terminal of the electric heating cable body. These connectors fusion-splice the single-mode fiber to an external standard single-mode fiber patch cord (LC / APC connector) for protection and achieve the same level of protection and sealing as the heating cable body, ensuring that the sensing signal is reliably led out of the fiber optic sensor demodulator. The intelligent slave unit integrates a passive intelligent slave unit during the manufacturing process of the electric heating cable at regular intervals (e.g., 30 meters). This module abandons the traditional wireless and independent power supply design, and innovatively adopts integrated power line carrier (PLC) communication and power supply technology. The module is directly connected to the power supply wire of the heating cable through a high-frequency coupling circuit, extracts the high-frequency carrier signal from the wire, and supplies power to its own circuit after rectification and voltage regulation; at the same time, this circuit is responsible for demodulating the commands issued by the master station and modulating and uploading data such as the equivalent impedance R_i(t) collected in the local area. The control center includes an edge gateway and an adaptive temperature control algorithm engine. The edge gateway, installed in the field control box, is responsible for driving the fiber optic sensor demodulator to acquire the temperature field and vibration signals of the entire line in real time; at the same time, it communicates with each intelligent slave unit to collect local electrical parameters (such as micro-area impedance); it is equipped with the following key interfaces: Fiber optic sensor demodulator interface: drives the demodulator to acquire the temperature field T(z,t) and vibration characteristics V(z,t) of the entire line in real time; Power line carrier master station modem: broadcasts control commands to all intelligent slave units via the heating cable itself, and receives data uploaded by each slave unit; Digital output module: used to control the relays of the zone power regulation unit. The adaptive temperature control algorithm engine runs within the edge gateway and incorporates multiple algorithm models proposed in this invention to perform data fusion, status analysis, heat demand prediction, and power decision-making. The zoned power regulation unit electrically divides the entire electric heating cable into several independent heating sections (corresponding to the location of the intelligent slave unit). Each heating section is connected to the main circuit of the control center through a solid-state relay (SSR) or intelligent power switch, and its on / off state and duty cycle receive independent commands from the control center. Electrical topology: adopts a topology of "AC total input → multiple independent branch outputs". After the main air switch, N independent circuits (N is the number of zones) are connected in parallel. Actuation device: each independent circuit consists of a solid-state relay (SSR), a branch protection circuit breaker, and a corresponding heating section of the heating cable connected in series. Control connection: the control terminal (low-voltage DC side) of each SSR is connected to the digital output module of the control center to receive the on / off (PWM) commands issued by it.

[0030] An adaptive temperature-regulating electric heating tape control method, employing an adaptive temperature-regulating electric heating tape system, includes the following steps: A1. Full-dimensional state perception and data fusion: The temperature distribution, vibration characteristics and equivalent impedance of the corresponding section are fused to construct a digital twin of the pipeline thermal state. Specifically, the edge gateway acquires the temperature distribution T(z,t) and vibration characteristics V(z,t) of the entire pipeline (1-meter spatial resolution) once per second using a fiber optic sensor demodulator. Simultaneously, the power line carrier (PLC) channel collects the segment equivalent impedance R_i(t) reported by each slave station. The adaptive temperature control algorithm engine integrates these three types of data with a known pipeline structure diagram (marked with valve and flange locations), a preset target temperature T_target, and the real-time ambient temperature T_env(t) acquired through the gateway to construct a digital twin of the pipeline's thermal state.

[0031] A2. Dynamic heat demand modeling and prediction: Run a dynamic heat flow balance and prediction model, and based on the current temperature, ambient temperature, and thermal resistance of the insulation layer, introduce a temperature change trend factor and a local disturbance factor to predict future heat demand. Specifically, real-time heat loss calculation: For each pipe unit (e.g., every 1 meter), the current heat dissipation rate is calculated based on its current temperature T(z,t), ambient temperature T_env(t), and dynamically updated total equivalent thermal resistance R_total(z,t). Q_loss(z,t)=(T(z,t)-T_env(t)) / R_total(z,t), where R_total(z,0)=R_total_init(z), whose initial value is calculated from the design drawing parameters. R_total(z,0) represents the total equivalent thermal resistance of the pipe at axial position z and time t=0 (initial moment), and R_total_init(z) represents the initial design calculation value of the total equivalent thermal resistance of the pipe at position z. It is a pre-calculated static array or function that is only related to position z.

[0032] R_total(z,0)=R_total_init(z) serves as a bridge connecting "design prior knowledge" and "data-driven learning." It ensures that the system starts operating from a reasonable, physics-based starting point, and then continuously self-corrects and optimizes through online data (T(z,t), T_env, P_in), causing the model parameter R_total(z,t) to gradually deviate from the initial theoretical value, ultimately approximating the actual thermal resistance of the pipeline under real, potentially altered (e.g., insulation dampness, aging) operating conditions.

[0033] Heat demand forecasting: Introducing a temperature change trend factor α(z,t)=[T(z,t)-T(z,t-Δt)] / Δt, where: T(z,t): the temperature value measured at position z at the end of the current control cycle t, T(z,t-Δt): the temperature value measured at position z at the end of the previous control cycle (t-Δt), Δt: the fixed control cycle of the system, i.e., the time interval between two consecutive executions of the control algorithm, [T(z,t)-T(z,t-Δt)]: the temperature change within one control cycle, and a local disturbance factor β(z,t) based on vibration analysis.

[0034] The predicted future heat demand is: the predicted linear power heat demand at position z and future time t+τ is: Q_demand_pred(z,t+τ)=Q_loss(z,t)*[1+γ1*α(z,t)+γ2*β(z,t)], where Q_loss(z,t) is the current heat dissipation rate, γ1 and γ2 are adaptive weight coefficients obtained through machine learning algorithm, and τ is the prediction step size.

[0035] For example, at a valve location, a sudden increase in the energy of a specific frequency band of V(z,t) may indicate a change in the medium flow rate, resulting in a corresponding increase in the β value and predicting an increase in heat demand at that point. The generation of β(z,t) involves first performing a short-time Fourier transform (STFT) on the vibration signal V(z,t) to obtain the time spectrum, and then calculating the signal energy E_band(z,t) of a specific frequency band (such as the band associated with fluid vortex shedding or valve throttling noise). β(z,t) can be obtained by normalizing the rate of change of this energy relative to the historical baseline energy, for example: β(z,t) = [E_band(z,t) - E_band_avg(z)] / E_band_avg(z), where E_band_avg(z) is the historical average energy at that point under steady state.

[0036] A3. Zonal adaptive power decision: Based on Q_demand_pred, calculate the required ideal heating power P_i_desired for each independent heating zone i. Specifically, the adaptive temperature control algorithm engine calculates the required ideal heating power P_i_desired for each independent heating segment i based on Q_demand_pred. Traditional PID control is replaced by a cooperative algorithm of model predictive control (MPC) and fuzzy rules. In step A3, based on the current state and the prediction model, the optimal power control sequence for each heating segment in the future finite time domain (prediction time domain Np) is solved and implemented in a rolling manner; Define variables: Control time domain: Nc (Nc≤Np), the control action usually remains unchanged after Nc steps.

[0037] State variables: , represents the total pipeline temperature distribution vector (consisting of discretization of T(z)) predicted for the next k steps (k=1,...,Np) at time t.

[0038] Control variables: , representing the power input vector of each heating segment for the next k steps (k=0,...,Nc-1) calculated at time t (M is the number of independent heating segments). The control variable u usually corresponds to the duty cycle of the SSR (between 0 and 1).

[0039] Output variable: y(k|t)=C*x(k|t), usually C=I, that is, the temperature is directly output.

[0040] Target setting: r, a vector with the same dimension as x, where each element is the target temperature T_target.

[0041] Predictive model (state-space form): x(k+1|t)=A*x(k|t)+B*u(k|t)+D*d(k|t)y(k|t)=C*x(k|t) A: The state transition matrix is ​​determined by the heat capacity of each unit in the pipeline, the thermal resistance between units, and the thermal resistance to convection (i.e., R_total), and is a large sparse matrix.

[0042] B: Control input matrix, describing how the power of each heating section affects the temperature of its associated pipe unit.

[0043] D: Perturbation input matrix.

[0044] d(k|t): Measurable perturbation vector, mainly containing the influence of T_env, which can be introduced through the feedforward of the Q_loss model in step A2.

[0045] Standard optimization problem: In each control cycle t, MPC solves the following constrained quadratic programming (QP) optimization problem: Minimize the objective function J(t): , in: The first term (tracking error term) penalizes the deviation between the predicted output and the setpoint. Q is a positive definite or semi-positive definite output error weight matrix. To improve uniformity, higher weights can be assigned to the diagonal elements corresponding to known "cold spots" (such as valves).

[0046] The second term (control increment term) penalizes drastic changes in the control quantity, smoothing the control action and avoiding impact on the SSR and the power grid. Δu(k|t)=u(k|t)-u(k-1|t), where R is the positive definite control increment weight matrix.

[0047] The third term (control energy term): directly penalizes the magnitude of the control quantity, reflecting the energy-saving target. S is the positive definite control quantity weight matrix, and ρ is the energy-saving preference coefficient.

[0048] The weighted L2 norm squared of vector v is denoted as . .

[0049] The following constraints must be met: System dynamics constraint: x(k+1|t)=A*x(k|t)+B*u(k|t)+D*d(k|t), for k=0,...,Np-1.

[0050] Output / State Constraints: T_min≤y(k|t)≤T_max, for k=1,...,Np (overheat protection and freeze protection).

[0051] Control input constraints: 0≤u(k|t)≤1, for k=0,...,Nc-1 (duty cycle upper and lower limits, 1 corresponds to 100% power).

[0052] Control increment constraint: |Δu(k|t)|≤Δu_max, for k=0,...,Nc-1 (limiting the rate of change of power per cycle).

[0053] Solution and Implementation: Solve the above QP problem online to obtain the optimal control sequence. The first element u*(0|t) of the sequence is taken as the actual control quantity (duty cycle command) applied to each heating section SSR in the current period t. The above process is repeated in the next period t+1 (rolling time-domain optimization).

[0054] Collaboration with fuzzy rules: When a fuzzy rule is triggered (such as "suspected insulation failure"), compensation can be achieved by dynamically adjusting the weight of the Q matrix at the corresponding "cold spot" position in the objective function, or by temporarily relaxing its T_min constraint, thereby achieving seamless collaboration with the MPC framework.

[0055] With the optimization objective of minimizing the temperature deviation between the entire section and the target temperature, and minimizing the total energy consumption, a constrained optimization problem is solved in each control cycle, and the optimal control sequence of power switching of each section in the next few cycles is directly output (the first sequence is executed). To address model uncertainties (such as changes in R_ins caused by moisture in the insulation layer), a set of fuzzy rules is designed. For example, if T(z) in a certain section is consistently lower than T_target AND the power command for that section has been consistently high AND its R_i(t) does not show any abnormalities, then the "suspected insulation failure" flag is triggered, and the calculation weight of P_i_desired for that section is temporarily and slightly increased; A4. Control commands are issued to the power regulation units of each zone for execution. At the same time, the system enters the feedback learning phase. If a systematic shift in the curve is detected, it is determined that the PTC material in that section has aged or undergone local microstructural changes. The power-temperature transfer function of that section is then corrected online during power decision-making.

[0056] Specifically, the feedback learning phase includes: Model parameter self-calibration: The actual measured temperature T(z,t+1) at the next time step is compared with the MPC model prediction T_pred(z,t+1) for the temperature at position z at time t+1 (i.e., at the end of the next control cycle). Using backpropagation of the error, the total equivalent thermal resistance R_total(z,t+1) of each pipe element at position z used for control calculations at time t+1 and thereafter is updated online using algorithms such as recursive least squares (RLS). The updated parameters are immediately used in step A2 of the next control cycle, forming a closed-loop learning process.

[0057] In step A4, the total equivalent thermal resistance R_total(z,t) of each discrete pipe element (position z) is dynamically updated so that the model-predicted temperature T_pred approximates the actual measured temperature T_meas. Model building: Within a short time window (such as several control cycles), assume that the thermal dynamics of pipe unit z is approximately a first-order system. Its discrete thermal equilibrium equation can be simplified to: T_meas(z,t)=a*T_meas(z,t-1)+b*P_in(z,t-1)+c*(T_env(t-1)-T_meas(z,t-1)), Where: P_in(z,t-1) is the thermal power input to this unit in the previous cycle (which can be calculated from the power distribution model of the tracing cable). T_env(t-1) is the ambient temperature in the previous cycle.

[0058] The parameter c is directly related to R_total, and c∝1 / R_total.

[0059] Rewrite the above equation in standard linear regression form: ,in: y(t)=T_meas(z,t) (system output). φ(t)=[T_meas (Observation data vector) (The parameter vector to be estimated, c is the core we are interested in). RLS Iteration Formula: For each pipe element z, in each control period t, after obtaining new observation data {y(t), φ(t)}, the parameter estimates θ_hat(t) are updated in the following order: Calculate the gain vector K(t): P(t) is the estimation error covariance matrix. The initial value P(0) is usually set to η*I (η is a large number, such as 1e5, and I is the identity matrix).

[0060] λ is the forgetting factor (0 < λ ≤ 1), usually taken as 0.95 to 0.999, used to make the algorithm pay more attention to new data, thereby tracking the time-varying parameter R_total.

[0061] Calculate the prior estimation error e(t): , Update parameter estimate θ_hat(t): θ_hat(t) = θ_hat(t-1) + K(t) * e(t), Update the covariance matrix P(t): , Thermal resistance extraction: Extract c_hat(t) from the updated parameter vector θ_hat(t). Based on the definition of thermal resistance and the system heat capacity C (which can be estimated from the pipe and medium properties and is a known constant), the dynamic thermal resistance estimate R_total_est(z,t) of this unit at the current moment can be calculated as: R_total_est(z,t) = (Δt) / (C * c_hat(t)), where Δt is the control period. This R_total_est(z,t) will be used as the updated value of R_total(z,t) and substituted into step A2 of the heat loss calculation in the next control period.

[0062] Online calibration of PTC characteristics: Analyze the relationship curve between R_i(t) and average temperature T_avg_i(t) for each segment and compare it with the theoretical RT curve of the PTC material. If a systematic shift in the curve is found, it is determined that the PTC material in that segment has aging or local microstructural changes, and the power-temperature transfer function of that segment is corrected online during power decision-making.

[0063] A computer medium storing a computer program that, when executed, implements the steps of an adaptive adjustment method for controlling an electric heating cable.

[0064] In this way, the "adaptability" of the system is ensured: the model is not fixed, but a dynamic entity that is constantly "polished" and "calibrated" by measured data as the system runs, thus enabling it to cope with various uncertainties and time-varying factors in real engineering environments.

[0065] It should be understood that the above embodiments are only for illustrating the technical concept and features of the present invention, and are intended to enable those skilled in the art to understand the content of the present invention and implement it accordingly. It should not be considered that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, several simple deductions or substitutions can be made without departing from the concept of the present invention. All equivalent changes or modifications made in accordance with the spirit and essence of the present invention should be covered within the protection scope of the present invention.

Claims

1. An adaptive temperature-regulating electric heating tape, comprising a liquid cooler, characterized in that: include: Electric heat tracing cable itself, The sensing unit includes a distributed temperature sensor, a distributed acoustic / vibration sensor, and a fiber optic sensing demodulator. The single-mode fiber of the fiber optic sensing demodulator is embedded between the insulation layer and the shielding layer along the axis of the electric heating cable body, serving as both a signal transmission channel and a sensing medium for the distributed temperature sensor and the distributed acoustic / vibration sensor. Fiber optic connectors are provided at the power supply cold end and the remote terminal of the electric heating cable body. These connectors fusion-splice the single-mode fiber to an external standard single-mode fiber jumper for reliable signal output from the fiber optic sensing demodulator. The intelligent slave unit integrates a passive intelligent slave unit into the electric heating cable body at regular intervals. It is directly connected to the power supply wire of the electric heating cable body through a high-frequency coupling circuit, extracts high-frequency carrier signals from the wire, and supplies power to its own circuit after rectification and voltage regulation. At the same time, it is responsible for demodulating the instructions issued by the power line carrier master station and modulating and uploading the data collected in the local area. The control center includes an edge gateway and an adaptive temperature control algorithm engine. The edge gateway, installed in the field control box, is responsible for driving the fiber optic sensor demodulator to acquire the temperature field and vibration signals of the entire line in real time; at the same time, it communicates with each intelligent slave unit to collect local electrical parameters; it is equipped with the following key interfaces: fiber optic sensor demodulator interface and driver demodulator interface to acquire the temperature field T(z,t) and vibration characteristics V(z,t) of the entire line in real time. Power line carrier master modem: broadcasts control commands to all intelligent slave units through the conductor of the electric heating cable itself, and receives data uploaded by each intelligent slave unit; Digital output module: Used to control the relays of the zone power regulation unit. The adaptive temperature control algorithm engine runs within the edge gateway, performing data fusion, status analysis, heat demand prediction, and power decision-making, including: Partition adaptive power decision: Based on Q_demand_pred, calculate the required ideal heating power P_i_desired for each independent heating zone i; Specifically, the adaptive temperature control algorithm engine calculates the required ideal heating power P_i_desired for each independent heating zone i based on Q_demand_pred. The traditional PID control is replaced by a collaborative algorithm of model predictive control and fuzzy rules. In step A3, based on the current state and the prediction model, the optimal power control sequence for each heating section within the future finite time domain is solved and implemented in a rolling manner. Define variables: Control time domain: Nc, the control action usually remains unchanged after Nc steps; State variables: , represents the predicted temperature distribution vector of the entire pipeline for the next k steps at time t; Control variables: , represents the power input vector of each heating section in the future k steps calculated at time t, and the control variable u usually corresponds to the duty cycle of the SSR; Output variable: y(k|t)=C*x(k|t), usually C=I, that is, the temperature is directly output; Target setting: r, a vector with the same dimension as x, where each element is the target temperature T_target; Prediction model: x(k+1|t)=A*x(k|t)+B*u(k|t)+D*d(k|t)y(k|t)=C*x(k|t) A: The state transition matrix is ​​determined by the heat capacity of each unit in the pipeline, the thermal resistance between units, and the thermal resistance to heat loss. It is a large sparse matrix. B: Control input matrix, describing how the power of each heating section affects the temperature of its associated pipe unit; D: Perturbation input matrix; d(k|t): Measurable perturbation vector, mainly containing the influence of T_env, which can be introduced through the Q_loss model feedforward in step A2; Standard optimization problem: In each control period t, MPC solves the following constrained quadratic optimization problem: Minimize the objective function J(t): , in: The first item is to penalize the deviation between the predicted output and the set value. Q is a positive definite or semi-positive definite output error weight matrix. In order to improve uniformity, higher weights can be assigned to the diagonal elements corresponding to the known "cold spots". The second term: punishes drastic changes in the control quantity to smooth the control action and avoid impacting the SSR and the power grid. Δu(k|t)=u(k|t)-u(k-1|t), where R is a positive definite control increment weight matrix. The third item: the magnitude of the direct penalty control quantity reflects the energy-saving target. S is the positive definite control quantity weight matrix, and ρ is the energy-saving preference coefficient. The weighted L2 norm squared of vector v is denoted as . ; The following constraints must be met: System dynamics constraint: x(k+1|t)=A*x(k|t)+B*u(k|t)+D*d(k|t), for k=0,...,Np-1; Output / state constraints: T_min≤y(k|t)≤T_max, for k=1,...,Np; Control input constraints: 0 ≤ u(k|t) ≤ 1, for k = 0, ..., Nc-1; Control increment constraint: |Δu(k|t)|≤Δu_max, for k=0,...,Nc-1; Solution and Implementation: Solve the above QP problem online to obtain the optimal control sequence. Only the first element u*(0|t) of the sequence is taken as the actual control quantity applied to each heating section SSR in the current period t, and the above process is repeated in the next period t+1. Collaboration with fuzzy rules: When a fuzzy rule is triggered, compensation can be achieved by dynamically adjusting the weights of the Q matrix at the corresponding cold point position in the objective function, or by temporarily relaxing its T_min constraint, thereby achieving seamless collaboration with the MPC framework. With the optimization objective of minimizing the temperature deviation between the entire heating zone and the target temperature, and minimizing the total energy consumption, a constrained optimization problem is solved in each control cycle, and the optimal control sequence of power switching of each heating zone in the next few cycles is directly output. To address model uncertainties, a set of fuzzy rules is designed. For example, if T(z) of a certain heating section is continuously lower than T_target AND the power command of the heating section has been continuously high AND its R_i(t) does not show any abnormality, then a suspected heat preservation failure flag is triggered, and the calculation weight of P_i_desired of the heating section is temporarily and slightly increased. Control commands are sent to the power regulation units of each zone for execution. At the same time, the system enters the feedback learning phase. If a systematic shift in the curve is detected, it is determined that the PTC material in the heating zone has aged or local microstructural changes. The power-temperature transfer function of the heating zone is then corrected online during power decision-making. The zoned power regulation unit electrically divides the entire electric heating cable into several independent heating sections. The heating sections correspond to the positions of the intelligent slave units. Each heating section is connected to the main circuit of the control center through a solid-state relay or intelligent power switch, and its on / off state and duty cycle receive independent commands from the control center.

2. A method for controlling an adaptive temperature-regulating electric heating tape, using the electric heating tape described in claim 1, characterized in that: include: A1. Full-dimensional state perception and data fusion: The temperature distribution, vibration characteristics and the equivalent impedance of the corresponding heating section are fused to construct a digital twin of the pipeline thermal state. A2. Dynamic heat demand modeling and prediction: Run a dynamic heat flow balance and prediction model, and based on the current temperature, ambient temperature, and thermal resistance of the insulation layer, introduce a temperature change trend factor and a local disturbance factor to predict future heat demand. A3. Zonal adaptive power decision: Based on the predicted heat demand power Q_demand_pred per unit length, calculate the required ideal heating power for each independent heating segment i. A4. Control commands are sent to the power adjustment units of each zone for execution. At the same time, the system enters the feedback learning phase. If a systematic shift in the curve is found, it is determined that the PTC material in the heating section has aged or local microstructure changes. The power-temperature transfer function of the heating section is corrected online during power decision-making. Partition adaptive power decision: Based on Q_demand_pred, calculate the required ideal heating power P_i_desired for each independent heating zone i; Specifically, the adaptive temperature control algorithm engine calculates the required ideal heating power P_i_desired for each independent heating zone i based on Q_demand_pred. The traditional PID control is replaced by a collaborative algorithm of model predictive control and fuzzy rules. In step A3, based on the current state and the prediction model, the optimal power control sequence for each heating section within the future finite time domain is solved and implemented in a rolling manner. Define variables: Control time domain: Nc, the control action usually remains unchanged after Nc steps; State variables: , represents the predicted temperature distribution vector of the entire pipeline for the next k steps at time t; Control variables: , represents the power input vector of each heating section in the future k steps calculated at time t, and the control variable u usually corresponds to the duty cycle of the SSR; Output variable: y(k|t)=C*x(k|t), usually C=I, that is, the temperature is directly output; Target setting: r, a vector with the same dimension as x, where each element is the target temperature T_target; Prediction model: x(k+1|t)=A*x(k|t)+B*u(k|t)+D*d(k|t)y(k|t)=C*x(k|t) A: The state transition matrix is ​​determined by the heat capacity of each unit in the pipeline, the thermal resistance between units, and the thermal resistance to heat loss. It is a large sparse matrix. B: Control input matrix, describing how the power of each heating section affects the temperature of its associated pipe unit; D: Perturbation input matrix; d(k|t): Measurable perturbation vector, mainly containing the influence of T_env, which can be introduced through the Q_loss model feedforward in step A2; Standard optimization problem: In each control cycle t, MPC solves the following constrained quadratic programming (QP) optimization problem: Minimize the objective function J(t): , in: The first item is to penalize the deviation between the predicted output and the set value. Q is a positive definite or semi-positive definite output error weight matrix. In order to improve uniformity, higher weights can be assigned to the diagonal elements corresponding to known cold spots. The second term: punishes drastic changes in the control quantity to smooth the control action and avoid impacting the SSR and the power grid. Δu(k|t)=u(k|t)-u(k-1|t), where R is a positive definite control increment weight matrix. The third item: the magnitude of the direct penalty control quantity reflects the energy-saving target. S is the positive definite control quantity weight matrix, and ρ is the energy-saving preference coefficient. The weighted L2 norm squared of vector v is denoted as . ; The following constraints must be met: System dynamics constraint: x(k+1|t)=A*x(k|t)+B*u(k|t)+D*d(k|t), for k=0,...,Np-1; Output / state constraints: T_min≤y(k|t)≤T_max, for k=1,...,Np; Control input constraints: 0 ≤ u(k|t) ≤ 1, for k = 0, ..., Nc-1; Control increment constraint: |Δu(k|t)|≤Δu_max, for k=0,...,Nc-1; Solution and Implementation: Solve the above QP problem online to obtain the optimal control sequence. Only the first element u*(0|t) of the sequence is taken as the actual control quantity applied to each heating section SSR in the current period t, and the above process is repeated in the next period t+1. Collaboration with fuzzy rules: When a fuzzy rule is triggered, compensation can be achieved by dynamically adjusting the weights of the Q matrix at the corresponding cold point position in the objective function, or by temporarily relaxing its T_min constraint, thereby achieving seamless collaboration with the MPC framework. With the optimization objective of minimizing the temperature deviation between the entire heating zone and the target temperature, and minimizing the total energy consumption, a constrained optimization problem is solved in each control cycle, and the optimal control sequence of power switching of each heating zone in the next few cycles is directly output. To address model uncertainties, a set of fuzzy rules is designed. For example, if T(z) of a certain heating section is continuously lower than T_target AND the power command of the heating section has been continuously high AND its R_i(t) does not show any abnormality, then the suspected heat preservation failure flag is triggered, and the calculation weight of P_i_desired of the heating section is temporarily and slightly increased. Control commands are sent to the power regulation units of each zone for execution. At the same time, the system enters the feedback learning phase. If a systematic shift in the curve is detected, it is determined that the PTC material in the heating zone has aged or undergone local microstructural changes. The power-temperature transfer function of the heating zone is then corrected online during power decision-making.

3. The adaptive temperature tracing heating cable control method according to claim 2, characterized in that: In step A1: the edge gateway acquires the temperature field T(z,t) and vibration characteristics V(z,t) of the entire pipeline once per second through the fiber optic sensor demodulator; the control center collects the equivalent impedance R_i(t) of the corresponding heating section reported by each intelligent slave unit through the power line carrier master station modem via the power line power supply conductor carrier channel of the electric heating cable. The adaptive temperature control algorithm engine integrates these three types of data with the known pipeline structure diagram, the preset target temperature T_target, and the real-time ambient temperature T_env(t) obtained through the edge gateway to construct a digital twin of the pipeline thermal state.

4. The adaptive adjustment and control method for electric heating tape according to claim 2, characterized in that: In step A2: specifically, real-time heat loss calculation: the current heat dissipation rate is calculated based on its current temperature T(z,t), ambient temperature T_env(t), and dynamically updated total equivalent thermal resistance R_total(z,t). Q_loss(z,t)=(T(z,t)-T_env(t)) / R_total(z,t), where R_total(z,0)=R_total_init(z), whose initial value is calculated from the design drawing parameters. R_total(z,0) represents the total equivalent thermal resistance of the pipe at axial position z and time t=0, and R_total_init(z) represents the initial design calculation value of the total equivalent thermal resistance of the pipe at position z; it is a pre-calculated static array or function that is only related to position z. Heat demand forecasting: Introducing a temperature change trend factor α(z,t)=[T(z,t)-T(z,t-Δt)] / Δt, where: T(z,t): current temperature, T(z,t-Δt): temperature value measured at position z at the end of the previous control cycle, Δt: fixed control cycle of the system, i.e., the time interval between two consecutive executions of the control algorithm, [T(z,t)-T(z,t-Δt)]: temperature change within one control cycle, and a local disturbance factor β(z,t) based on vibration analysis. The predicted linear power heat demand at position z and future time t+τ is: Q_demand_pred(z,t+τ)=Q_loss(z,t)*[1+γ1*α(z,t)+γ2*β(z,t)], where Q_loss(z,t) is the current heat dissipation rate, γ1 and γ2 are adaptive weight coefficients obtained through machine learning algorithm, and τ is the prediction step size.

5. The adaptive temperature tracing heating cable control method according to claim 2, characterized in that: In step A4: specifically, the feedback learning phase includes: Model parameter self-calibration: The actual measured temperature T(z,t+1) at the next time step is compared with the MPC model prediction value T_pred(z,t+1) for the temperature of position z at the next time step t+1 at time t. Using backpropagation of the error, the total equivalent thermal resistance R_total(z,t+1) of each pipe unit at position z used for control calculations at time t+1 and thereafter is updated online through algorithms such as recursive least squares. The updated parameters are immediately used in step A2 of the next control cycle, forming a closed-loop learning. Online calibration of PTC characteristics: Analyze the relationship curve between R_i(t) and average temperature T_avg_i(t) for each heating section and compare it with the theoretical RT curve of PTC material; if a systematic shift in the curve is found, it is determined that the PTC material in that heating section has aging or local microstructural changes, and the power-temperature transfer function of that heating section is corrected online during power decision.

6. A computer medium having a computer program stored thereon, characterized in that: When the program is executed, it implements the steps of the method as described in any one of claims 2-5.

Images