High-precision active global temperature maintaining system for static leveling instrument

Abstract

The invention relates to the field of engineering structure health monitoring, and discloses a high-precision active global temperature maintaining system for a static level gauge, which comprises a temperature sensing network for fusing a quasi-distributed fiber bragg grating array and a discrete digital temperature sensor and performing high-precision real-time monitoring on a global temperature field; the bidirectional active heat flow control unit adopts a semiconductor chilling plate array to carry out active and bidirectional (refrigeration / heating) precise temperature regulation on any point of the system; the intelligent control core runs a three-level intelligent control algorithm based on machine learning prediction, Kalman filtering data fusion and fuzzy adaptive PID adjustment, processes sensing data and generates an accurate temperature control instruction; all the components of the system are integrated in a multi-layer composite thermal management sheath with aerogel as a core thermal insulation layer. According to the invention, a uniform and constant temperature field is created and maintained on the physical level, so that measurement errors caused by temperature gradient are eradicated, and the monitoring precision and reliability are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of health monitoring of engineering structures, and more particularly to a high-precision active global temperature maintenance system for hydrostatic levels. Background Technology

[0002] A hydrostatic leveling system (HLS) is a high-precision relative elevation measurement instrument based on the principle of communicating vessels in classical fluid statics. The system consists of a series of probes (or reservoirs) interconnected by liquid-filled communicating vessels. When the system is stationary, under the influence of gravity, the liquid surfaces within all probes lie on the same equipotential surface. By deploying a high-precision level sensor (such as a capacitive, inductive, or vibrating wire sensor) within each probe, the distance from the sensor's reference plane to the liquid surface can be accurately measured. By comparing the level readings of different probes, the relative elevation changes between monitoring points can be precisely calculated.

[0003] Due to their high precision, automation, and long-term stability, hydrostatic levels are widely used in the health monitoring of large and super-large engineering structures. For example, they are used to monitor key parameters such as bridge deflection and pier settlement, dam deformation, uneven settlement of high-rise buildings, and ground and adjacent building settlement caused by tunnel construction, providing crucial data support for engineering safety assessment and maintenance.

[0004] Although the principle of a hydrostatic level is simple, its measurement accuracy is easily affected by external environmental factors, especially temperature changes, in practical engineering applications. Even though some commercial products claim to have built-in temperature compensation functions, the actual results are often unsatisfactory. Temperature is the most critical and complex factor affecting the accuracy of HLS measurements, and its influence mechanism is mainly reflected in the following aspects.

[0005] Liquid density variation: This is the primary source of error. According to the fundamental hydrostatic formula P=ρgh, the physical quantity measured by the sensor at the measuring point is the pressure exerted by the liquid, and the liquid level is calculated from this pressure. Liquid density is a key parameter in this formula. However, liquid density changes with temperature. When the connecting pipes of an HLS system are exposed to a non-uniform temperature field, the liquid temperature in different pipe sections will differ, causing the density to no longer be a constant throughout the system. In this case, even if the physical liquid level is exactly the same at all probes, the pressure at the bottom of each measuring point will vary due to local density differences, ultimately leading to significant errors in the calculated elevation reading.

[0006] Thermal expansion and contraction of system components: All physical components of the HLS system, including metal or plastic reservoirs, sensor housings, and connecting pipes, are subject to the physical law of thermal expansion and contraction. When the ambient temperature changes, the geometry of these components will change. For example, changes in the diameter and length of the connecting pipe will alter the total system volume, and the expansion or contraction of the sensor mounting base will directly change the height of the measurement reference point. These minute changes in physical dimensions will accumulate and ultimately be reflected in the error of the elevation reading.

[0007] Temperature gradient effect: In large-scale structural monitoring, HLS pipelines may stretch for tens or even hundreds of meters. Some pipelines may be exposed to direct sunlight, while others may be in the shade of buildings or buried underground. This complex on-site environment inevitably places the system in a dynamic, non-uniform temperature field, i.e., a significant temperature gradient exists. The temperature gradient is the fundamental cause of the two error mechanisms mentioned above, and also the most important and complex reason for HLS measurement errors.

[0008] To address the aforementioned thermally induced error problem, the engineering and academic communities have explored various technical approaches, but all of them have significant limitations.

[0009] Table 1 Comparative Analysis of Existing Thermal Management Technologies

[0010] For the reasons mentioned above, the market urgently needs a high-precision active global temperature maintenance system for hydrostatic levels to keep the hydrostatic levels operating under optimal testing conditions. Summary of the Invention

[0011] The present invention aims to at least solve one of the technical problems existing in the related art. To this end, the present invention provides a high-precision active global temperature maintenance system for hydrostatic levels.

[0012] A high-precision active global temperature maintenance system for hydrostatic levels. A temperature sensing network is used to monitor the temperature distribution of the hydrostatic level in real time to generate temperature monitoring data. The intelligent control core is connected to the temperature sensing network, receives the temperature monitoring data, and generates bidirectional temperature control commands based on the temperature monitoring data and the preset target temperature. A bidirectional active heat flow control unit, connected to the intelligent control core, is used to receive the bidirectional temperature control command and actively adjust the heating or cooling of the hydrostatic level according to the command.

[0013] Furthermore, the bidirectional active heat flow control unit includes a plurality of semiconductor cooling chips distributed along the hydrostatic level.

[0014] 3. Further, the temperature sensing network is a hybrid sensing network, comprising: A quasi-distributed fiber Bragg grating temperature sensing array laid along the connecting conduit of the hydrostatic level; and A discrete digital temperature sensor network deployed in the probe or valve area of ​​the hydrostatic level.

[0015] Furthermore, the intelligent control core also includes performing data fusion processing on the temperature monitoring data received from the hybrid sensor network to generate a global temperature field estimate. Based on the deviation between the estimated global temperature field and the preset target temperature and the rate of change of the deviation, the bidirectional temperature control command is generated through a fuzzy adaptive PID control algorithm.

[0016] Furthermore, the data fusion processing employs a Kalman filter-based model to fuse the heterogeneous and noisy data from the hybrid sensor network into the global temperature field estimate.

[0017] Furthermore, the intelligent control core also includes, Using a machine learning-based predictive model, future temperature change trends can be predicted based on historical temperature data and historical control outputs. A feedforward control signal is generated based on the predicted temperature change trend, and it is superimposed with the feedback control signal generated by the fuzzy adaptive PID control algorithm to form the final bidirectional temperature control command.

[0018] Furthermore, it also includes, A multi-layer composite thermal management sheath encloses the hydrostatic level, the temperature sensing network, and the bidirectional active heat flow control unit; the sheath includes a core insulation layer made of aerogel insulation felt.

[0019] The above-described one or more technical solutions in the embodiments of the present invention have at least one of the following technical effects: The core design principle of this scheme is to completely shift from the traditional approach of "passive adaptation and post-compensation" to "active intervention and source elimination." Its goal is to create and maintain a uniform and constant temperature field directly at the physical level by constructing a closed-loop, intelligent, active thermal management system, thereby eradicating HLS measurement errors caused by temperature gradients.

[0020] 1) Multi-layer composite thermal management jacket, using ultra-low thermal conductivity aerogel insulation felt as the core insulation layer, and integrating a bidirectional active heat flow control unit; 2) Bidirectional active heat flow control, using a high-reliability thermoelectric cooler (TEC) instead of a single heating film to achieve active, bidirectional (cooling / heating) precise temperature regulation at any point in the system; 3) A hybrid high-density distributed sensor network integrates a quasi-distributed fiber Bragg grating temperature sensor array with a discrete high-precision digital temperature sensor to achieve real-time monitoring of the temperature field across the entire area with high density and high precision. The control core employs a high-performance microcontroller, running a three-level intelligent control algorithm based on machine learning prediction, fuzzy adaptive PID regulation, and Kalman filter data fusion.

[0021] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention are described clearly and completely below. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention. The following embodiments are used to illustrate this invention but cannot be used to limit the scope of this invention.

[0023] The technical solution of this invention mainly relates to a high-precision active global temperature maintenance system for hydrostatic leveling systems (HLS). This system aims to fundamentally eliminate measurement errors caused by temperature gradients and fluctuations, thereby improving the monitoring accuracy and reliability of HLS.

[0024] I. System Overall Architecture This system is an integrated system that combines advanced materials, sensing technology, thermal management technology and intelligent control algorithms.

[0025] Its overall architecture can be divided into four core subsystems, which work together: Multi-layer composite thermal management jacket: Serving as the system's "exoskeleton and skin," it wraps around the entire HLS system, providing efficient thermal insulation, mechanical protection, and integrating functional components.

[0026] Bidirectional active heat flow control actuator: As the "muscle" of the system, it is composed of an array of semiconductor cooling chips and is responsible for precisely heating or cooling any part of the HLS system according to instructions.

[0027] Distributed temperature sensing network: As the "neural network" of the system, it is composed of a combination of fiber optic gratings and digital temperature sensors to monitor the temperature distribution across the entire system in real time and with high density.

[0028] Intelligent control and decision-making: As the "brain" of the system, it consists of a high-performance microcontroller and advanced algorithms running on it, and is responsible for processing sensor data, making decisions and directing the execution units to work.

[0029] The advanced nature of this solution stems from the organic synergy and deep integration of multiple cutting-edge technology modules. The selection of each technology precisely maps to and addresses one or more specific pain points in existing technologies.

[0030] 1. Multi-layer composite thermal management jacket This jacket is the physical basis for the system's thermal management and environmental protection. Its design aims to provide superior thermal insulation, reliable mechanical protection, uniform heat conduction, and integrated support for internal functional components. It employs a modular, flexible design to accommodate the complex piping and equipment shapes of HLS systems. Its structure consists of three layers from the outside in: Outer protective and mechanical support layer: Utilizing high-strength, weather-resistant, and UV-resistant flexible composite materials, such as PTFE-coated fiberglass cloth. This layer provides robust external protection for the entire sheath, resisting scratches, impacts at the construction site, and long-term exposure to wind and rain, ensuring the safety of the internal functional layers. This meets the high durability requirements of geotechnical engineering monitoring instruments in harsh field environments.

[0031] Core insulation layer: Aerogel insulation felt: This is the key to achieving efficient thermal management.

[0032] Traditional insulation materials (such as rock wool and rubber-plastic composites) have relatively high thermal conductivity, resulting in limited insulation performance. While vacuum insulation panels (VIPs) have extremely low thermal conductivity, they are rigid panels, making them difficult to wrap irregularly shaped HLS pipes, valves, and joints. Furthermore, their vacuum structure completely loses its insulation performance if accidentally punctured, leading to poor reliability. In contrast, aerogel insulation felt (such as Aspen Aerogels' Pyrogel® XTE series) is currently the best overall performance option. It not only possesses extremely low thermal conductivity comparable to VIPs (its thermal conductivity at room temperature can be lower than...), but also... Furthermore, it possesses exceptional flexibility, easily wrapping any complex geometric shape like a regular blanket. Its extremely low thermal conductivity means that, to achieve the same insulation effect, its required thickness is only 1 / 3 to 1 / 5 of traditional materials. This significantly reduces the size and weight of the entire sheathing system, facilitating installation in space-constrained areas. Secondly, it exhibits excellent hydrophobicity, effectively preventing moisture intrusion, which is crucial for maintaining long-term stable insulation performance and protecting internal electronic components. Finally, it demonstrates good physical robustness and strong compressive strength, recovering its thermal properties even after being subjected to pressure.

[0033] Inner Functional Integration Layer: This layer is in close contact with the surface of the controlled HLS device and serves as the carrier for functional components. It utilizes a flexible printed circuit board (FPC) or a similar flexible substrate. The TEC active thermal control unit, the DS18B20 digital temperature sensor, and their connecting circuitry are integrated or bonded to this layer. The material of this layer must possess good thermal conductivity to ensure that the heat generated by the TEC can be efficiently and evenly transferred to the surface of the HLS system.

[0034] 2. Bidirectional active heat flow control actuator: Semiconductor refrigeration technology To address the fundamental flaw of unidirectional heating in the initial concept, this solution employs a thermoelectric cooler (TEC), also known as a Peltier module, as the core actuator for active heat flow control.

[0035] The TEC module operates based on the Peltier effect. When direct current passes through a thermocouple made of two different semiconductor materials, one end absorbs heat (cools), while the other releases heat. Crucially, by changing the direction of the direct current, the cooling and heating sides can be switched instantaneously. This unique bidirectional heat pump characteristic perfectly solves the complex needs of HLS systems, which require both heating (e.g., maintaining temperature in cold environments) and cooling (e.g., eliminating localized overheating under sunlight), forming the technological cornerstone for achieving true "constant temperature maintenance."

[0036] 3. Hybrid high-density distributed temperature sensing network Any single sensor technology involves compromises in terms of cost, accuracy, ease of deployment, and environmental adaptability. To achieve optimized temperature field monitoring of the complex HLS system (which includes long, slender pipes and a structured, concentrated sensor head), this solution innovatively adopts a hybrid sensing strategy that organically combines the advantages of quasi-distributed fiber optic sensing and discrete digital sensing.

[0037] 3.1 Quasi-distributed sensing - Fiber Bragg grating (FBG) array: It is mainly used to monitor long-distance connecting liquid pipe sections in HLS systems.

[0038] FBG sensing technology boasts a series of unparalleled advantages. First, as an optical sensing technology, it is inherently immune to electromagnetic interference (EMI), making it ideal for use in industrial environments with strong electromagnetic interference, such as power plants and near high-voltage lines. Second, FBG sensors are small and lightweight, allowing for easy integration into a sheath. Most importantly, it possesses powerful multiplexing capabilities: dozens or even hundreds of FBG sensors with different reflection wavelengths can be inscribed on the same optical fiber, enabling long-distance "quasi-distributed" measurements via a single fiber, greatly simplifying wiring.

[0039] Technical Principle and Implementation: A periodic refractive index modulation is formed in the fiber core using ultraviolet or femtosecond lasers, creating a grating. This grating reflects light of a specific wavelength, the Bragg wavelength. When the external temperature changes, the Bragg wavelength shifts due to the thermo-optical effect (changing the effective refractive index) and the thermal expansion effect (changing the grating period). By monitoring the wavelength shift using a high-precision fiber Bragg grating demodulator, the temperature change can be accurately determined. Its temperature sensitivity is typically around 10 pm / °C, and when combined with a demodulator with picometer (pm) resolution, a temperature measurement resolution of 0.1°C or even higher can be achieved.

[0040] 3.2 Discrete Node Sensing - Single-Bus Digital Temperature Sensor (DS18B20) Network: It is mainly used for monitoring areas with relatively complex structures and concentrated dimensions, such as HLS measuring heads, valves, and controller modules. In these areas, high-density multi-point temperature monitoring is required at a low cost.

[0041] The DS18B20 is a high-performance digital temperature sensor with excellent cost-effectiveness. Its core advantage lies in its single-wire protocol, allowing multiple sensors to be connected in parallel on the same data line. Only a single microcontroller GPIO pin is needed to communicate with the entire sensor network, significantly simplifying wiring complexity and reducing costs in localized areas. Its output is a digital signal, eliminating the need for an external ADC. Within the common engineering temperature range of -10°C to +85°C, it achieves an accuracy of ±0.5°C, with a resolution of up to 0.0625°C in 12-bit mode, fully meeting the requirements of this system's control algorithm. Furthermore, its unique parasitic power supply mode allows it to draw power solely from the data line, further simplifying power supply design and making it particularly suitable for space-constrained applications.

[0042] 4. Core of Intelligent Control and Decision-Making: Embedded Systems and Advanced Algorithms The system's "brain" is an embedded system based on a high-performance microcontroller, responsible for running complex sensing, fusion, and control algorithms.

[0043] Microcontroller (MCU): This solution uses the STM32H7 series high-performance microcontroller from STMicroelectronics, such as the STM32H743ZI.

[0044] Control Strategy Overview: This scheme abandons simple switching control or traditional fixed-parameter PID control, and proposes an innovative three-level hybrid intelligent control strategy to cope with the complex characteristics of the system such as nonlinearity, time-varying nature and thermal inertia.

[0045] Bottom layer: Fuzzy adaptive PID feedback control. Responsible for real-time, robust closed-loop temperature regulation.

[0046] Middle layer: Multi-sensor data fusion based on Kalman filtering. It is responsible for fusing the heterogeneous and noisy data from the hybrid sensor network into a unified, smooth, and more realistic global temperature field estimate, providing high-quality input for top-level and bottom-level control.

[0047] Top layer: Predictive feedforward control based on machine learning. It is responsible for learning the thermodynamic behavior of the system and the patterns of external disturbances (such as the solar cycle), predicting future temperature change trends, and performing compensatory control in advance to proactively overcome the thermal inertia of the system.

[0048] This layered and integrated advanced control architecture is the core innovation that distinguishes this solution from existing temperature control systems (such as those using only fuzzy PID). While solving old problems (such as temperature gradients), it also anticipates and addresses new challenges arising from the introduction of new technologies (such as multi-source data fusion and overcoming thermal inertia). The resolution of these new challenges demonstrates the technological depth and completeness of this solution.

[0049] The core of this solution lies in its hierarchical and integrated control algorithm, which integrates traditional PID control, modern fuzzy logic, state estimation algorithm and cutting-edge machine learning technology.

[0050] 4.1 Design of Fuzzy Adaptive PID Control Algorithm This is the system's underlying feedback controller, responsible for real-time, robust temperature regulation.

[0051] Basic PID Controller: The core algorithm uses incremental PID. Unlike positional PID, which directly calculates the control output, incremental PID calculates the increment of the control output. The formula is: It is the increment of the control quantity at the k-th sampling time; , These are the control quantities at the k-th and (k-1)th sampling times, respectively; , , These are the proportional coefficient, integral coefficient, and differential coefficient, respectively. , , These are the deviation values ​​at the k-th, (k-1), and (k-2)th sampling times, respectively. The advantage of incremental algorithms is that they do not include past error accumulation in the calculation, so the calculation error has a smaller impact on the control quantity, and there will be no serious impact when the controller switches from manual to automatic.

[0052] Introducing fuzzy logic for parameter adaptation: The performance of traditional PID controllers is highly dependent on... , and Tuning of three parameters. For a complex thermal system like HLS, which has nonlinear, large inertia, and time-varying characteristics, a fixed set of PID parameters is unlikely to achieve optimal performance under all operating conditions. Therefore, this scheme introduces fuzzy logic to perform online, intelligent, and adaptive adjustment of the PID parameters.

[0053] Fuzzy controller design: Establish a fuzzy controller with temperature error E (Error) and error change rate EC (ErrorChange) as dual inputs, and PID parameter adjustment (…). , , () is a three-output fuzzy controller.

[0054] Fuzzification: The precise numerical values ​​input (e.g., E = +2.5°C, EC = -0.1°C / s) are mapped to fuzzy linguistic variables through a membership function. For example, the fuzzy set of E and EC can be defined as {NB (negative large), NS (negative small), ZO (zero), PS (positive small), PB (positive large)}.

[0055] Fuzzy rule base: This is the core of fuzzy control's "expert knowledge." It defines the relationship between inputs and outputs through a series of "IF-THEN" rules. For example, a rule used to adjust... The rule could be: "IF E is PB (positive) ANDEC is ZO (zero), THEN..." "is PB (Zhengda)". The physical meaning of this rule is: if the temperature deviation is large and tends to stabilize, it indicates that the system response is too slow and the proportional term needs to be increased significantly. To speed up response times, the entire rule base can be visually represented using a two-dimensional table.

[0056] Defuzzification: After fuzzy inference, the resulting fuzzy output needs to be converted into clear, executable PID parameter adjustment values ​​using a defuzzification method (such as the centroid method). These adjustment values ​​will update the PID controller in real time. , , .

[0057] Table 2: Example of a fuzzy logic control rule base (for adjustment) )

[0058] (Note: PB - positive large, PM - positive medium, PS - positive small, ZO - zero, NS - negative small, NM - negative medium, NB - negative large) Anti-Windup and Bumpless Transfer: In practical control, when the output power of the TEC module reaches its physical upper limit (100%) or lower limit (0%), the controller output saturates. At this time, if the standard PID integral term continues to accumulate error, it will cause the integrator to "over-saturate" or "windup". When the system error reverses and the controller needs to respond, this over-saturated integral term takes a long time to "unwind", resulting in severe system response lag and huge overshoot. Similarly, when the controller switches from manual mode to automatic mode, or when the setpoint changes significantly, the PID output will jump instantaneously if left untreated, causing a shock to the system, i.e., "bump".

[0059] This solution integrates anti-integral saturation and disturbance-free switching mechanisms to address the above issues.

[0060] Anti-saturation: A back-calculation strategy is employed. This method involves the controller calculating the output... Compared with the actual saturated actuator output Compare, difference Multiplied by a feedback gain This feedback is sent back to the input of the integrator, thus "holding" the integrator in place and preventing it from accumulating excessively.

[0061] Bumpy switching: At the instant of switching from manual to automatic mode, the initial value of the integral term of the PID controller is forcibly set to the current manual output value, or the controller output is smoothly transitioned from a manual value to an automatically calculated value through tracking mode, thereby avoiding abrupt output changes. These complex control logics are modeled, simulated, and verified in the Simulink environment to ensure robustness under various boundary conditions.

[0062] 4.2 Multi-sensor data fusion model based on Kalman filtering Hybrid sensor networks provide the system with massive, redundant, yet heterogeneous temperature data. To provide the PID controller with a unified, smooth, and more accurate global temperature state estimate, it is essential to effectively fuse this multi-source heterogeneous data. This solution employs a Kalman filter to achieve this goal.

[0063] Establish a state-space model of the temperature field: Apply Kalman filtering, first describe the behavior of the system using a set of state-space equations.

[0064] State transition equation (prediction model): Observation equations (measurement model): in: : The system state vector at time k. In this scheme, it includes the temperature T and the rate of temperature change of all n control nodes of the system. : State transition matrix, describing how the system state naturally evolves from the previous time k−1 to the current time k.

[0065] : Control input model matrix, which describes how the control input affects the state changes.

[0066] : The control input vector at time k−1, i.e., the power applied to each TEC unit.

[0067] : Process noise vector, representing the uncertainty of the model itself.

[0068] : The observation vector at time k, i.e., the actual readings of all m sensors.

[0069] The observation matrix represents the true state vector. Linearly mapped to the sensor's reading space.

[0070] The measurement noise vector represents the inaccuracy of the sensor's measurements. Implementation process: Within each control cycle, the MCU will execute a complete Kalman filter iteration: Prediction step: Using the state transition equation, estimate the optimal state from the previous time step. and control input Predict the state at the current moment. and its uncertainty (covariance matrix) ).

[0071] Update step: Read all sensor measurements Calculate the Kalman gain. This gain balances the reliability of the predicted and measured values. Finally, the measured values ​​are used to correct the predicted state, resulting in the optimal state estimate for the current time step. and the updated covariance matrix .

[0072] 4.3 Predictive Temperature Feedforward Control Based on Machine Learning HLS systems exhibit significant thermal inertia, and traditional feedback control is essentially "retroactive" control. To achieve ultimate temperature stability and rapid response to external disturbances (such as sunlight), feedforward control must be introduced, which "predicts" future disturbances and takes compensatory actions in advance.

[0073] Model selection and training: This approach uses a nonlinear autoregressive external input (NARX) neural network or a similar recurrent neural network (RNN) variant (such as LSTM) as the prediction model.

[0074] Feedforward control implementation: In each control cycle, the trained ML model uses input features such as historical temperature, historical control output, and environmental parameters to predict the temperature drift that will occur in the system.

[0075] Based on this prediction, the feedforward controller calculates in reverse the cooling or heating power that needs to be applied in advance to offset this temperature rise.

[0076] This feedforward control signal will be superimposed with the feedback control signal calculated by the fuzzy adaptive PID feedback controller to form the final TEC drive signal.

[0077] The training and deployment of machine learning models are already quite mature, so only a brief explanation is given here.

[0078] Training phase: Data acquisition for model training. During the initial system installation, a "data acquisition mode" is run to record system temperature, environmental parameters, and control output under different weather and seasons to generate training samples.

[0079] Model deployment and updates: The model can be pre-trained offline and then embedded in the microcontroller, or it can support online learning or remote firmware updates (OTA) to optimize the model according to seasonal changes or changes in the field environment.

[0080] Specific implementation methods 1. Packaging process and on-site installation procedure for thermal management sleeves Encapsulation: The sheath is modularly prefabricated at the factory according to the standard pipe diameter and probe size of the HLS system. The excellent flexibility of the aerogel insulation felt makes it very easy to cut and shape. The TEC module and DS18B20 sensor network are pre-integrated into the inner functional layer of the sheath at the factory and brought out with a unified electrical interface for "plug and play".

[0081] Installation process: Surface preparation: Ensure that the surfaces of HLS equipment and piping are clean, dry, and free of oil to guarantee good thermal contact.

[0082] Wrapping and securing: Wrap the prefabricated sheath around the corresponding HLS component and secure it tightly with stainless steel cable ties or special tape.

[0083] Joint treatment: The joints between sheath modules are potential "thermal bridges" and should be connected by overlapping and sealed with special aerogel sealing tape or sealant.

[0084] 2. Packaging, protection, and installation of FBG sensors Bare optical fibers are very fragile and must be robustly encapsulated and protected.

[0085] Packaging solution: Probe and equipment surface: The FBG sensor is packaged using a solderable or adhesive method.

[0086] Connecting conduits: The FBG sensor array is threaded through miniature stainless steel tubing or armored fiber optic cable to provide superior mechanical protection.

[0087] Temperature and strain cross-sensitivity decoupling: The Bragg wavelength shift of the FBG is simultaneously affected by both temperature and strain, as shown in the following equation: in: Bragg wavelength; The amount of Bragg wavelength shift; : The axial strain change experienced by the optical fiber; Temperature change; : The effective elastic-optical coefficient of optical fiber; : Coefficient of thermal expansion of optical fiber; Thermo-optic coefficient of optical fiber.

[0088] In this scheme, to eliminate the impact of strain crosstalk on temperature measurement accuracy, strain isolation encapsulation is mainly adopted, so that the optical fiber is in a free and relaxed state without force inside the encapsulation tube, thereby physically isolating the transmission of external strain.

[0089] In light of the limitations of current mainstream thermal management technologies in the market and academia, the advantages of the technical solution of this invention will be explained.

[0090] Passive insulation technology The most common method is to wrap the HLS piping and probes with traditional insulation materials (such as foam, glass wool, rubber, etc.). This passive insulation aims to slow down the rate of heat exchange between the system and the external environment.

[0091] Limitations: Passive insulation can only provide temporary relief, not a fundamental solution. It cannot actively eliminate existing temperature differences, nor can it resist the internal temperature gradient caused by the continuous action of strong external heat sources such as sunlight. In long-term monitoring environments with large temperature differences, heat will eventually penetrate the insulation layer, leading to temperature imbalance within the system. Therefore, its effectiveness is very limited and far from meeting the ideal requirement of "isothermal conditions across the entire area" under high-precision monitoring.

[0092] Active heating compensation technology The core idea is to lay heating wires or films inside the insulation layer and monitor the temperature of different parts of the system using temperature sensors. When the controller detects temperature inconsistencies, it activates heating to raise the temperature of the colder parts, aiming to achieve temperature balance throughout the system.

[0093] Limitations: This method has a fundamental flaw—it's unidirectional control. It can only heat up, not cool down. In many real-world scenarios, such as under strong summer sunlight, the temperature of some pipes can be far higher than other parts of the system and the set target temperature. In this situation, simple heating will be ineffective in cooling the overheated parts, thus failing to achieve true temperature stability and balance.

[0094] Challenges of Software Algorithm Compensation Another mainstream approach involves installing temperature sensors at each measuring point of the HLS, and then using software algorithms to perform post-correction on the collected elevation data. This method attempts to fit the relationship between temperature and error using a mathematical model.

[0095] Limitations: The failure of software compensation stems not from the algorithm itself, but from a severe lack of physical layer information. First, for continuous pipes stretching tens of meters or even longer, installing only a few temperature sensors at the measuring points is completely insufficient to capture the complex, dynamically changing distributed temperature field along the pipe. This is akin to attempting to describe the shape of a complex curve using a few points, inevitably leading to severe distortion of the model's input information. Second, establishing a universally applicable and accurate temperature-error model is extremely difficult because it requires comprehensively considering the thermophysical properties (density, coefficient of thermal expansion, etc.) of various materials such as the liquid, pipe material, and sensor housing at different temperatures. These parameters are inherently difficult to obtain precisely. Therefore, any attempt to perform mathematical fitting under incomplete information will be significantly less effective, which is the fundamental reason for the poor performance of existing manufacturer-manufactured compensation solutions.

[0096] The core design principle of this scheme is to completely shift from the traditional approach of "passive adaptation and post-compensation" to "active intervention and source elimination." Its goal is to create and maintain a uniform and constant temperature field directly at the physical level by constructing a closed-loop, intelligent, active thermal management system, thereby eradicating HLS measurement errors caused by temperature gradients.

[0097] The core of the solution lies in the following innovations: 1) Multi-layer composite thermal management jacket, using ultra-low thermal conductivity aerogel insulation felt as the core insulation layer, and integrating a bidirectional active heat flow control unit; 2) Bidirectional active heat flow control, using a high-reliability thermoelectric cooler (TEC) instead of a single heating film to achieve active, bidirectional (cooling / heating) precise temperature regulation at any point in the system; 3) A hybrid high-density distributed sensor network integrates a quasi-distributed fiber Bragg grating temperature sensor array with a discrete high-precision digital temperature sensor to achieve real-time monitoring of the temperature field across the entire area with high density and high precision. The control core employs a high-performance microcontroller, running a three-level intelligent control algorithm based on machine learning prediction, fuzzy adaptive PID regulation, and Kalman filter data fusion.

[0098] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A high-precision active global temperature maintenance system for a hydrostatic level, characterized in that, A temperature sensing network is used to monitor the temperature distribution of the hydrostatic level in real time to generate temperature monitoring data. The intelligent control core is connected to the temperature sensing network, receives the temperature monitoring data, and generates bidirectional temperature control commands based on the temperature monitoring data and the preset target temperature. A bidirectional active heat flow control unit, connected to the intelligent control core, is used to receive the bidirectional temperature control command and actively adjust the heating or cooling of the hydrostatic level according to the command.

2. The high-precision active global temperature maintenance system for a hydrostatic level according to claim 1, characterized in that, The bidirectional active heat flow control unit includes multiple semiconductor cooling chips distributed along the hydrostatic level.

3. The high-precision active global temperature maintenance system for a hydrostatic level according to claim 1 or 2, characterized in that, The temperature sensing network is a hybrid sensing network, comprising: A quasi-distributed fiber Bragg grating temperature sensing array laid along the connecting conduit of the hydrostatic level; and A discrete digital temperature sensor network deployed in the probe or valve area of ​​the hydrostatic level.

4. The high-precision active global temperature maintenance system for a hydrostatic level according to claim 3, characterized in that, The intelligent control core also includes performing data fusion processing on the temperature monitoring data received from the hybrid sensor network to generate a global temperature field estimate. Based on the deviation between the estimated global temperature field and the preset target temperature and the rate of change of the deviation, the bidirectional temperature control command is generated through a fuzzy adaptive PID control algorithm.

5. The high-precision active global temperature maintenance system for a hydrostatic level according to claim 4, characterized in that, The data fusion processing is performed using a Kalman filter-based model to fuse the heterogeneous and noisy data from the hybrid sensor network into the global temperature field estimate.

6. The high-precision active global temperature maintenance system for a hydrostatic level according to claim 4 or 5, characterized in that, The intelligent control core also includes, Using a machine learning-based predictive model, future temperature change trends can be predicted based on historical temperature data and historical control outputs. A feedforward control signal is generated based on the predicted temperature change trend, and it is superimposed with the feedback control signal generated by the fuzzy adaptive PID control algorithm to form the final bidirectional temperature control command.

7. The high-precision active global temperature maintenance system for a hydrostatic level according to claim 1, characterized in that, It also includes, A multi-layer composite thermal management sheath encloses the hydrostatic level, the temperature sensing network, and the bidirectional active heat flow control unit; the sheath includes a core insulation layer made of aerogel insulation felt.

Images