Optical fiber-palladium alloy membrane monohydrogen sensor based on temperature compensation and control method of optical fiber-palladium alloy membrane monohydrogen sensor

Abstract

The invention discloses an optical fiber-palladium alloy membrane single hydrogen sensor machine control method based on temperature compensation, and relates to the technical field of online monitoring of power equipment. A hydrogen-sensitive sensing unit and a temperature compensation module are arranged in the sensor. The hydrogen-sensitive sensing unit and the temperature compensation module are arranged in parallel, are completely immersed in insulating oil, are in contact with oil flow and are used for monitoring dissolved hydrogen in situ in real time; the hydrogen-sensitive sensing unit adopts an optical fiber structure of which the side wall is plated with a palladium alloy nano-film, and is used for sensing the concentration of hydrogen dissolved in oil and generating corresponding optical signal change; the temperature compensation module adopts an optical fiber structure inert to hydrogen and is used for synchronously measuring the oil temperature of the position where the probe is located. By means of the structure and the matched control method, in-situ and real-time monitoring of the concentration of the hydrogen in the sleeve is achieved, the reliability and stability of hydrogen concentration monitoring data are greatly improved, a more accurate basis is provided for early fault diagnosis of a transformer, and the device and the method have great significance in guaranteeing safe operation of a power grid.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of online monitoring and fault diagnosis of power equipment, and in particular to a fiber-palladium alloy film single hydrogen sensor based on temperature compensation and a control method thereof. BACKGROUND

[0002] High-voltage oil-filled bushings are key outgoing line components of large power equipment such as transformers and reactors, and the internal insulation performance thereof is directly related to the operation safety of the entire substation. During operation, the bushing will cause the internal insulation oil to decompose and produce hydrogen due to conductor heating, dielectric loss, and possible partial discharge or overheating faults. Hydrogen is the most initial and typical characteristic gas of electric thermal decomposition of insulating materials, and its rapid growth is an important early indicator of internal latent faults of the bushing.

[0003] At present, the monitoring of oil-filled bushings relies mainly on periodic power-off maintenance and oil chromatography tests, and it is difficult to achieve early detection and early warning of faults. Existing online monitoring technologies are difficult to directly and safely implant sensors into the narrow space filled with insulation oil inside the bushing. Even if it can be installed, traditional electrochemical or semiconductor gas sensors have the problems of poor long-term stability, the need for regular calibration, and serious temperature cross-sensitivity. The internal oil temperature of the bushing changes dramatically with load and environment, and the temperature change not only affects the solubility of hydrogen in the oil, but also directly causes the characteristics of the sensitive elements of the sensor to drift, causing measurement data distortion, and thus causing system false alarms or missed alarms, which cannot provide reliable basis for operation and maintenance decisions. SUMMARY

[0004] The purpose of the present application is to provide a fiber-palladium alloy film single hydrogen sensor based on temperature compensation and a control method thereof, which can adapt to the complex oil-immersed environment inside the bushing and realize accurate and reliable online monitoring of the dissolved hydrogen concentration in the insulation oil through real-time temperature compensation.

[0005] In order to realize the above technical problems, the purpose of the present application is to provide a temperature compensation-based optical fiber-palladium alloy film single hydrogen sensor. The sensor is provided with a hydrogen-sensitive sensing unit, a temperature compensation module, a pressure balancing unit, an optical coupler, a thermal management unit, a power management module, a signal demodulation host, a temperature compensation processor, a data fusion unit and a communication interface; the probe assembly is externally sealed in a protective shell. The sensor includes a miniaturized probe assembly packaged in a miniature metal probe rod, which is implanted in the oil cavity of the transformer oil-filled bushing; the probe assembly includes a hydrogen-sensitive sensing unit and a temperature compensation module packaged in parallel, and the hydrogen-sensitive sensing unit and the temperature compensation module are completely immersed in the insulating oil and in contact with the oil flow for in-situ real-time monitoring of dissolved hydrogen; the hydrogen-sensitive sensing unit adopts an optical fiber structure with a palladium alloy nanometer film plated on the side wall for sensing the dissolved hydrogen concentration in the oil and generating corresponding optical signal changes; the temperature compensation module adopts a hydrogen-inert optical fiber structure for synchronous measurement of the oil temperature at the position of the probe.

[0006] The hydrogen-sensitive sensing unit captures the original information of the dissolved hydrogen concentration in the oil. When hydrogen interacts with the sensitive film, it causes changes in its optical properties, thereby generating a mixed original response containing the true hydrogen signal and the temperature interference signal; the temperature compensation module senses the instantaneous oil temperature at the position of the probe, which remains inert to external hydrogen changes, and its output is a pure temperature signal; the physical proximity and environmental consistency of the hydrogen-sensitive sensing unit and the temperature compensation module ensure the synchronization of the collected data in space and time.

[0007] The hydrogen-sensitive sensing unit uses a multilayer composite structure, and the palladium alloy nanometer film sensor is designed with a gradient thickness, and its hydrogen-sensitive response characteristics satisfy the following nonlinear equation:

[0008] In the formula, , , is the temperature-dependent response coefficient, is the linear compensation factor, is the hydrogen partial pressure.

[0009] The temperature measurement model of the temperature compensation module is:

[0010] wherein, , is the temperature sensitivity coefficient matrix element, is the wavelength drift of the jth reference light.

[0011] The pressure balancing unit and thermal management unit serve a protective function. The pressure balancing unit, through its internal flexible structure, dynamically balances the oil pressure inside and outside the probe, preventing minor deformation of the probe structure due to pressure changes and avoiding the introduction of additional, non-hydrogen-sensitive effects that cause optical signal drift, thus stabilizing the internal mechanical environment of the probe assembly. The thermal management unit, through a heat conduction path, conducts away or balances the heat generated by internal components or drastic changes in external oil temperature, reducing the temperature gradient inside the probe and preventing damage to optical components or measurement errors caused by local overheating or thermal stress.

[0012] The optical coupler is located at the signal output end of the hydrogen-sensitive sensing unit and the temperature compensation module. It couples two independent optical signals from the hydrogen-sensitive sensing unit and the temperature compensation module into the same transmission optical fiber. The coupled composite optical signal passes through the fiber outlet of the sealed protective shell and enters the oil-resistant and high-voltage-resistant single-mode optical cable for transmission to the distant end.

[0013] The signal demodulation host emits broadband probe light into the optical cable and receives the composite optical signal carrying hydrogen and temperature information reflected back from the probe end. The high-resolution spectral analysis system inside the signal demodulation host is responsible for decoding the composite optical signal, separating and calculating the wavelength shift of the hydrogen-sensitive sensing unit and the wavelength shift of the temperature compensation module, and converting them from optical quantities into digital quantities.

[0014] The temperature compensation processor receives the real-time data stream from the signal demodulation host, removes the interference caused purely by temperature effects from the hydrogen-sensitive signal, corrects the sensor sensitivity for changes in temperature, and finally outputs the absolute hydrogen concentration value.

[0015] The data fusion unit receives the current absolute hydrogen concentration value output by the temperature compensation processor, combines it with historical data trends and temperature change rate information within the data fusion unit, assesses the evolution trend of the equipment status, performs a preliminary diagnosis of the sensor's own health status, and converts the internal data format and related status information confirmed by the data fusion unit into a standardized industrial communication protocol via the communication interface 10. This allows for interface communication with the substation's integrated automation system, status monitoring master station, or cloud monitoring platform, enabling remote monitoring, historical recording, and alarm functions.

[0016] A control method for a temperature-compensated fiber-optic palladium alloy film single hydrogen sensor: Step 1: Information is collected through the hydrogen-sensitive sensing unit and temperature compensation module. The Kalman filter module performs recursive optimal estimation on the noisy wavelength drift signal collected by the hydrogen-sensitive sensing unit. Through its continuous prediction and update steps, random interference is dynamically suppressed and the signal-to-noise ratio is improved. Step 2: The pre-processed clean hydrogen-sensitive signal and the real-time temperature data provided by the temperature compensation module are sent into the model of the temperature compensation processor, which outputs a fully temperature-compensated and accurate absolute hydrogen concentration value, completing the intelligent conversion from noisy raw data to reliable monitoring results. Step 3: The output of the temperature compensation processor is fused by the data fusion unit and then output through the communication structure. Compared with existing technologies, the beneficial effects of this invention are as follows: This invention has inherent advantages such as intrinsic safety and strong resistance to electromagnetic interference. Through an innovative intelligent temperature compensation mechanism, it effectively eliminates the influence of oil temperature fluctuations, and achieves high-precision and high-stability in-situ monitoring of hydrogen concentration, providing a reliable basis for early fault diagnosis of transformers. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein: Figure 1 This is a circuit block diagram of the present invention; Figure 2 This is a flowchart of the signal processing and temperature compensation algorithm of the present invention. Detailed Implementation

[0018] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0019] A temperature-compensated fiber-optic palladium alloy film single-hydrogen sensor is disclosed. The core of the sensor is a miniaturized probe assembly encapsulated within a micro-metal probe rod. Its dimensions are precisely designed for implantation into the oil cavity via a pre-installed oil valve or a dedicated mounting interface. This probe assembly integrates a hydrogen-sensitive sensing unit 1 and a temperature compensation module 2, both packaged in parallel. The hydrogen-sensitive sensing unit 1 uses a fiber grating or micro / nano fiber with a nano-palladium alloy thin film coated on its sidewalls as the sensing core. When hydrogen interacts with the palladium film, it causes changes in the optical refractive index and mechanical strain, resulting in a characteristic drift in the grating wavelength or optical phase. The temperature compensation module 2 uses a hydrogen-inert but temperature-sensitive fiber grating. Its function is to synchronously and in-situ measure the precise oil temperature at the probe's location, providing a reference signal for subsequent compensation.

[0020] Wavelength shift of the hydrogen-sensitive unit: = f (hydrogen concentration, temperature) Wavelength shift of the temperature-compensated unit: = f (temperature) Let the total wavelength shift of the hydrogen-sensitive fiber unit be... It is the partial pressure of hydrogen. With ambient temperature A bivariate function can be expressed as:

[0021] in: This indicates that the hydrogen-sensitive unit is at a certain temperature. The sensitivity coefficient is itself a function of temperature.

[0022] It is a function of hydrogen partial pressure and describes the intrinsic relationship between hydrogen concentration and optical response.

[0023] This represents the wavelength shift caused by the hydrogen-sensitive unit itself due to pure temperature effects (such as thermal expansion and thermo-optic effects). Meanwhile, the output of the temperature-compensated fiber unit is only affected by the ambient temperature. Sensitive, its wavelength shift It can be represented as:

[0024] in It is the temperature sensitivity coefficient of the temperature compensation unit. For reference temperature, This is the initial offset. By simultaneously solving the above equations and utilizing the precise temperature value provided by the temperature compensation unit... (Depend on (obtained through real-time inversion), the system can obtain the total response from the hydrogen-sensitive unit. Accurately subtract the pure temperature component Finally, the intrinsic hydrogen-sensitive response is determined by querying or calculating the temperature-corrected value. And combine Henry's law to determine the partial pressure of hydrogen gas. Converting to the dissolved concentration in oil will output a true and accurate hydrogen concentration value that has undergone full temperature compensation.

[0025] The host unit's embedded dedicated algorithm can synchronously demodulate the wavelength changes of the hydrogen sensor and the temperature compensation unit, and drive an intelligent temperature compensation module. This module, as the core software algorithm, uses real-time acquired temperature data to correct and decouple the original response signal of the hydrogen sensor in real time, ultimately outputting an accurate absolute hydrogen concentration value unaffected by temperature. Furthermore, the probe assembly is encapsulated entirely in inorganic materials to ensure the chemical stability and mechanical reliability required for long-term operation in high-temperature insulating oil.

[0026] Specifically, such as Figure 1As shown, the sensor includes a hydrogen sensing unit 1, a temperature compensation module 2, a pressure balancing unit 3, an optical coupler 4, a thermal management unit 5, a power management module 6, a signal demodulation host 7, a temperature compensation processor 8, a data fusion unit 9, and a communication interface 10.

[0027] The system begins with a miniature probe assembly immersed in insulating oil. Hydrogen-sensitive sensing unit 1 and temperature compensation module 2 are encapsulated in parallel within the probe, both exposed to the same oil immersion environment. Hydrogen-sensitive sensing unit 1 directly captures the raw information of dissolved hydrogen concentration in the oil. When hydrogen interacts with the sensitive membrane, it causes changes in its optical properties, generating a mixed raw response containing both the actual hydrogen signal and temperature interference signals. Almost simultaneously, the adjacent temperature compensation module 2 focuses on accurately sensing the instantaneous oil temperature of the probe. It remains inert to changes in external hydrogen levels, and its output is a pure temperature signal. The physical proximity and environmental consistency of these two modules ensure that the data they acquire are highly synchronized and comparable in time and space, laying the fundamental foundation for subsequent accurate temperature compensation.

[0028] However, the interior of a transformer bushing presents a challenging environment, where fluctuations in oil pressure and uneven temperature distribution can threaten sensitive measurements. This is where the pressure balancing unit 3 and thermal management unit 4 play crucial roles. The pressure balancing unit 3, through its flexible internal structure, dynamically balances the oil pressure inside and outside the probe, preventing minor deformations of the probe structure due to pressure changes. This avoids introducing additional, non-hydrogen-sensitive effects that could cause optical signal drift, ensuring the hydrogen-sensitive and temperature-compensated modules operate in a relatively stable mechanical environment. Meanwhile, the thermal management unit 5 works to establish a more uniform and stable thermal field inside the probe. Through optimized heat conduction paths, it quickly removes or equalizes heat generated by internal components or sudden changes in external oil temperature, minimizing the temperature gradient inside the probe and preventing damage to optical components or measurement errors caused by localized overheating or thermal stress.

[0029] Once the hydrogen sensing and temperature signals are successfully captured and stabilized, they need to be integrated and transmitted remotely via optical coupler 4. Optical coupler 4 serves as the bridge connecting the probe assembly to the rear section. Through a precise micro-optical structure, it efficiently and with low loss couples the two independent optical signals from the hydrogen sensing unit 1 and the temperature compensation module 2 into the same transmission optical fiber. This process not only achieves signal multiplexing but also ensures that the polarization state and intensity of the optical signals remain as stable as possible during coupling, providing a high-quality input signal for subsequent precise demodulation. The coupled composite optical signal then passes through the fiber exit hole of the sealed protective housing into an oil-resistant and high-voltage-resistant single-mode optical cable, beginning its transmission to the remote end.

[0030] The signal demodulation host 7 is the "signal conversion center" of the entire system. It first transmits broadband probe light into the optical cable and receives the composite optical signal carrying hydrogen and temperature information reflected back from the probe end. The host's internal high-resolution spectral analysis system is responsible for "decoding" this composite signal. Through complex algorithms, it accurately separates and calculates the wavelength shift corresponding to the hydrogen-sensitive sensing unit 1 and the temperature compensation module 2, converting them from optical quantities to digital quantities. During this process, the host's built-in advanced filtering algorithm also smooths and denoises the original digital signal to suppress various random noises introduced during transmission and detection, refining a purer, more effective signal.

[0031] At this point, the system obtains the pre-processed hydrogen-sensitive digital signal and temperature digital signal, which are then fed into the temperature compensation processor 8 in real time. The temperature compensation processor 8 is the core component for achieving high-precision measurement. Internally, it runs a pre-trained neural network model based on an improved particle swarm optimization algorithm. This model deeply understands the complex nonlinear mapping relationship between hydrogen concentration, ambient temperature, and the final output of the hydrogen-sensitive sensing unit 1. The processor receives the real-time data stream from the demodulation host 7 and, using this intelligent model, dynamically and in real-time extracts the interference purely caused by temperature effects from the hydrogen-sensitive signal and corrects for changes in sensor sensitivity with temperature. After this series of complex intelligent calculations, the processor finally outputs a highly accurate absolute hydrogen concentration value.

[0032] The precise concentration value is processed by the data fusion unit 9. This unit also incorporates historical data trends and temperature change rates over a period of time for more advanced data fusion and inference, thereby assessing the evolution of the equipment's status, providing a preliminary diagnosis of the sensor's health, and further improving the reliability of the output data. The final result confirmed by the data fusion unit 9—the precise hydrogen concentration and related status information—is then converted into a standardized industrial communication protocol by the communication interface 10. This allows for interface communication with the substation's integrated automation system, status monitoring master station, or cloud monitoring platform, enabling remote monitoring, historical recording, and alarm functions. The power management module 6 draws power from the industrial power supply and converts, regulates, and isolates it to various voltage and power levels required by the signal demodulation host 7 and temperature compensation processor 8 of each module.

[0033] Hydrogen-sensitive sensing unit 1 uses a multilayer composite structure, and the palladium alloy nanofilm sensing employs a gradient thickness design. Its hydrogen-sensitive response characteristics satisfy the following nonlinear equation:

[0034] In the formula, , , The temperature-dependent response coefficient, It is a linear compensation factor. This is the partial pressure of hydrogen gas.

[0035] The temperature measurement model of temperature compensation module 2 is as follows:

[0036] in, , These are the elements of the temperature sensitivity coefficient matrix. Let be the wavelength shift of the j-th reference light subfield.

[0037] Pressure balancing unit 3, its pressure compensation function is:

[0038] in, , This is the pressure compensation coefficient. For real-time hydraulic pressure, For reference pressure.

[0039] Optical coupler 4 consists of a multiplexing unit, an optical intensity stabilizer, and a polarization controller, enabling precise coupling and transmission of optical signals.

[0040] Optical coupler 4 enables precise coupling and transmission of multiple optical signals. Its transmission characteristics satisfy:

[0041] in, For the k-th level coupling efficiency, The polarization-dependent loss coefficient, This is the phase modulation amount.

[0042] Thermal management unit 5 ensures the sensor's measurement stability under drastic temperature changes. Its thermal balance equation is:

[0043] in, Let Q be the thermal diffusivity, and Q be the heat source term. is the convective heat transfer coefficient.

[0044] The signal demodulation host 7 employs a multi-stage filtering strategy in its signal processing flow.

[0045] in, Let be the transfer function of the i-th stage filter. This represents the Fourier transform.

[0046] The temperature compensation processor 8 employs an improved particle swarm optimization-backpropagation neural network algorithm, with the network training objective function being:

[0047] in, For network parameters, The regularization coefficient is . denoted as the number of neurons in the i-th layer.

[0048] Data fusion unit 9 achieves deep fusion of multi-sensor information and uses D-S evidence theory for uncertainty reasoning.

[0049] in, Assign values ​​to the basic probability of proposition A. For different sources of evidence, the confidence function is denoted as .

[0050] like Figure 2 The diagram shows a flowchart of a signal processing and temperature compensation algorithm for a fiber optic single-hydrogen sensor. It details the core temperature compensation algorithm of this invention, namely the complete workflow of a hybrid intelligent model based on Kalman filtering and improved particle swarm optimization neural network (PSO-BP).

[0051] The process begins with the preprocessing of the original optical signal. The Kalman filter module first performs recursive optimal estimation on the noisy wavelength drift signal acquired by the hydrogen-sensitive unit. Through its continuous prediction and update steps, it dynamically suppresses random interference, significantly improving the signal-to-noise ratio and providing a clean input source for subsequent compensation. The preprocessed clean hydrogen-sensitive signal and the real-time temperature data provided by the temperature compensation unit are then fed into the core calculation loop of the PSO-BP neural network compensation module. This module first prepares the data and initializes the network, and then initiates the improved particle swarm optimization process: the algorithm encodes the weights and thresholds of the neural network into the spatial positions of the particle swarm, evaluates the quality by calculating the fitness of the parameter set represented by each particle, and dynamically updates the individual historical best position and the global best position. In this process, the algorithm adopts a dual adaptive strategy, adjusting the inertia weight according to the iteration quality and decreasing the learning factor over time, thereby balancing global exploration and local exploitation capabilities and effectively overcoming the premature convergence problem.

[0052] Once the particle swarm optimization meets the termination condition, the optimal initial parameters are assigned to the backpropagation (BP) neural network. The network then enters a supervised fine-tuning training phase, where parameters are precisely calibrated through backpropagation of errors. Ultimately, a high-precision nonlinear mapping model is established from "temperature" and "filtered hydrogen-sensitive signal" to "actual hydrogen concentration." In the real-time monitoring phase, the system directly inputs the real-time collected temperature and hydrogen-sensitive signals into this fully trained model, instantly outputting a precise absolute hydrogen concentration value with full temperature compensation, thus completing the intelligent conversion from noisy raw data to reliable monitoring results.

Claims

1. A temperature-compensated fiber-optic palladium alloy membrane single-hydrogen sensor, characterized in that: The sensor includes a miniaturized probe assembly encapsulated within a miniature metal probe rod, which is implanted in the oil cavity of the transformer's oil-filled bushing. The probe assembly includes a hydrogen-sensitive sensing unit (1) and a temperature compensation module (2) encapsulated in parallel. The hydrogen-sensitive sensing unit (1) and the temperature compensation module (2) are completely immersed in insulating oil and in contact with the oil flow for in-situ real-time monitoring of dissolved hydrogen. The hydrogen-sensitive sensing unit (1) adopts an optical fiber structure with a palladium alloy nanofilm coated on the sidewalls for sensing the concentration of dissolved hydrogen in the oil and generating corresponding optical signal changes. The temperature compensation module (2) adopts an optical fiber structure inert to hydrogen for synchronously measuring the oil temperature at the probe's location.

2. The temperature-compensated fiber-palladium alloy membrane single hydrogen sensor according to claim 1, characterized in that: The sensor is equipped with a pressure balancing unit (3), an optical coupler (4), a thermal management unit (5), a power management module (6), a signal demodulation host (7), a temperature compensation processor (8), a data fusion unit (9), and a communication interface (10); the probe assembly is enclosed in a sealed protective housing.

3. The temperature-compensated fiber-palladium alloy membrane single hydrogen sensor according to claim 1, characterized in that: The hydrogen-sensitive sensing unit (1) captures the original information of dissolved hydrogen concentration in the oil. When hydrogen interacts with the sensitive membrane, it causes a change in its optical properties, thereby generating a mixed original response containing real hydrogen signals and temperature interference signals. The temperature compensation module (2) senses the instantaneous oil temperature at the probe location. It remains inert to changes in external hydrogen and its output is a pure temperature signal. The physical proximity and environmental consistency of the hydrogen-sensitive sensing unit (1) and the temperature compensation module (2) ensure that the collected data are synchronized in time and space.

4. A temperature-compensated fiber-palladium alloy membrane single-hydrogen sensor according to claim 3, characterized in that: The hydrogen-sensitive sensing unit (1) uses a multilayer composite structure, and the palladium alloy nanofilm sensing adopts a gradient thickness design. Its hydrogen-sensitive response characteristics satisfy the following nonlinear equation: In the formula, , , The temperature-dependent response coefficient, It is a linear compensation factor. This is the partial pressure of hydrogen gas. The temperature measurement model of the temperature compensation module (2) is as follows: in, , These are the elements of the temperature sensitivity coefficient matrix. Let be the wavelength shift of the j-th reference light subfield.

5. A temperature-compensated fiber-palladium alloy membrane single-hydrogen sensor according to claim 2, characterized in that: The pressure balancing unit (3) and the thermal management unit (5) serve a protective function. The pressure balancing unit (3) uses an internal flexible structure to dynamically balance the oil pressure inside and outside the probe, preventing slight deformation of the probe structure due to pressure changes, avoiding the introduction of additional, non-hydrogen-sensitive effects that cause optical signal drift, and ensuring the stability of the mechanical environment inside the probe assembly. The thermal management unit (5) uses a heat conduction path to conduct away or balance the heat generated by internal components or drastic changes in external oil temperature, reducing the temperature gradient inside the probe and preventing damage to optical components or measurement errors caused by local overheating or thermal stress.

6. A temperature-compensated fiber-palladium alloy membrane single-hydrogen sensor according to claim 2, characterized in that: The optical coupler (4) is set at the signal output end of the hydrogen-sensitive sensing unit (1) and the temperature compensation module (2) to couple two independent optical signals from the hydrogen-sensitive sensing unit (1) and the temperature compensation module (2) into the same transmission optical fiber. The coupled composite optical signal enters the oil-resistant and high-pressure-resistant single-mode optical cable through the fiber outlet hole of the sealed protective shell and is transmitted to the far end.

7. A temperature-compensated fiber-palladium alloy membrane single-hydrogen sensor according to claim 2, characterized in that: The signal demodulation host (7) emits broadband detection light into the optical cable and receives the composite optical signal carrying hydrogen and temperature information reflected back from the probe end. The high-resolution spectral analysis system inside the signal demodulation host (7) is responsible for decoding the composite optical signal, separating and calculating the wavelength shift of the hydrogen-sensitive sensing unit (1) and the wavelength shift of the temperature compensation module (2), and converting them from optical quantities to digital quantities.

8. A temperature-compensated fiber-palladium alloy membrane single-hydrogen sensor according to claim 2, characterized in that: The temperature compensation processor (8) receives the real-time data stream from the signal demodulation host (7), removes the interference caused purely by temperature effect from the hydrogen-sensitive signal, corrects the sensor sensitivity as temperature changes, and finally outputs the absolute hydrogen concentration value.

9. A temperature-compensated fiber-optic palladium alloy membrane single-hydrogen sensor according to claim 2, characterized in that: The data fusion unit (9) receives the current absolute hydrogen concentration value output by the temperature compensation processor (8), combines the historical data trend and temperature change rate information in the data fusion unit (9), evaluates the evolution trend of the equipment status, performs a preliminary diagnosis of the health status of the sensor itself, and converts the internal data format into a standardized industrial communication protocol through the communication interface 10 to connect with the substation's integrated automation system, status monitoring master station or cloud monitoring platform to realize remote monitoring, historical record and alarm functions of the data.

10. A control method for a temperature-compensated fiber-optic palladium alloy membrane single-hydrogen sensor according to any one of claims 1-9, characterized in that: Step 1: Information is collected through the hydrogen-sensitive sensing unit (1) and the temperature compensation module (2). The Kalman filter module performs recursive optimal estimation on the noisy wavelength drift signal collected by the hydrogen-sensitive sensing unit (1). Through its continuous prediction and update steps, random interference is dynamically suppressed and the signal-to-noise ratio is improved. Step 2: The pre-processed clean hydrogen-sensitive signal and the real-time temperature data provided by the temperature compensation module (2) are sent into the model of the temperature compensation processor (8) to output a fully temperature-compensated and accurate absolute hydrogen concentration value, thus completing the intelligent conversion from noisy raw data to reliable monitoring results. Step 3: The output of the temperature compensation processor (8) is fused by the data fusion unit (9) and then output to the outside through the communication structure (10).

Images