Multi-spectral radiation temperature measurement method and system for improving quantum particle swarm
By introducing Tent chaotic mapping and Lévy flight mechanism to optimize the particle swarm optimization algorithm, the problems of local extrema and poor adaptability in multispectral radiometric temperature measurement are solved, achieving higher temperature measurement accuracy and robustness, and making it suitable for multispectral radiometric temperature measurement in complex environments.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HARBIN ENGINEERING UNIVERSITY SANYA NANHAI INNOVATION & DEVELOPMENT BASE
- Filing Date
- 2026-04-20
- Publication Date
- 2026-05-19
AI Technical Summary
Existing particle swarm optimization algorithms are prone to getting trapped in local extrema and have poor adaptability in multispectral radiation thermometry. They struggle to find the global optimum in complex high-dimensional and nonlinear search spaces, resulting in low temperature measurement accuracy.
Tent chaotic mapping is introduced to optimize particle initialization and the Lévy flight mechanism. The initial particle position is generated by Tent chaotic mapping and updated by Lévy flight mechanism, which enhances global search capability and adaptive adjustment. The particle position is updated by adaptive contraction-expansion coefficient and standard normal distribution random number, so as to achieve fast and accurate inversion of temperature and spectral emissivity.
It improves the inversion performance of multispectral radiation thermometry, enhances robustness and temperature measurement accuracy under complex working conditions, and is suitable for non-destructive and accurate measurement in high temperature, ultra-high temperature and inaccessible harsh environments.
Smart Images
Figure CN122062801A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the fields of radiation thermometry and photoelectric signal processing technology, specifically to an improved multispectral radiation thermometry method and system for quantum particle swarm optimization. Background Technology
[0002] Multispectral radiation thermometry is an advanced non-contact temperature measurement technology. Its basic principle is to simultaneously acquire radiation signals from the target object in multiple discrete wavelength bands, and based on Planck's blackbody radiation law, establish a physical model relating the radiation signal, true temperature, and spectral emissivity. Finally, the true temperature of the target is estimated by solving an inverse problem. Compared to traditional contact temperature measurement methods, multispectral radiation thermometry effectively overcomes inherent limitations such as thermal disturbance, response delay, and sensor loss due to its unique advantages of being non-contact, having a fast response speed, a wide temperature measurement range, and not disrupting the surface temperature field of the target object. This technology is particularly suitable for high-temperature, ultra-high-temperature, and inaccessible harsh environments, enabling non-destructive and accurate measurement of transient and dynamic temperature fields. It has already been maturely applied in key fields such as metallurgy, aerospace, industrial heating, and energy monitoring.
[0003] While this technology avoids the problems of thermal disturbance, response hysteresis, and sensor loss inherent in contact temperature measurement, its core scientific challenge lies in the coupling between temperature and spectral emissivity. This is because the spectral emissivity of any real object... All of these are unknown and vary complexly with factors such as wavelength, temperature, surface condition, and observation angle, leading to... Inversion of measurements from each spectral channel One unknown (1 true temperature value + 1 unknown) The process of calculating emissivity values at various wavelengths constitutes an inherently underdetermined and ill-conditioned inverse problem. To solve this underdetermined problem in multispectral radiometric thermometry, traditional methods generally rely on pre-defined emissivity-wavelength models (such as linear, logarithmic, or polynomial relationships), transforming the problem into a nonlinear fitting based on that model; or establishing empirical emissivity-temperature relationships. However, the accuracy of these inversions is highly dependent on the correctness of the pre-defined model and has poor universality. Furthermore, emerging deep learning technologies enhance flexibility through data-driven feature learning, but are constrained by their reliance on massive amounts of accurate data and high computational costs. In conclusion, in complex industrial environments, factors such as oxidation, contamination, and roughness variations on material surfaces lead to extremely complex emissivity behavior, which is difficult to describe accurately with simple mathematical models. Any model deviation will be directly transmitted and amplified into the final temperature inversion error, severely limiting the reliability and application scope of traditional methods.
[0004] In recent years, intelligent optimization algorithms have provided new approaches to solving this inverse problem. These algorithms transform the temperature and emissivity inversion process into a constrained optimization problem, searching for the global optimum by minimizing the residual between the measured spectrum and the theoretically calculated spectrum, thereby reducing the dependence on explicit emissivity models. Among them, Particle Swarm Optimization (PSO) has attracted much attention due to its simple concept, few parameters, and ease of implementation. However, when applying the standard PSO algorithm to high-dimensional, nonlinear multispectral inversion, its limitations become increasingly apparent: 1. Prone to getting trapped in local extrema: Due to the extremely complex nonlinear coupling between spectral emissivity and wavelength, traditional methods have an imbalance between global exploration and local exploitation capabilities, lack long-distance jump capabilities, and particles are prone to clustering or leaving blind spots in the solution space. Once they fall into the local extremum region, they are difficult to escape and cannot find the optimal solution that approximates the true temperature in the global range, which seriously affects the temperature measurement accuracy.
[0005] 2. Poor adaptability: When faced with the complex search space of high-dimensional and nonlinear multispectral inversion, it is difficult to adapt to different stages of the entire optimization process, lacks an adaptive adjustment mechanism, and has weak robustness.
[0006] Therefore, developing a robust improved optimization algorithm with stronger search capabilities and higher convergence accuracy is of great research significance and engineering value for improving the inversion performance of multispectral radiometric temperature measurement technology under complex working conditions and promoting its wider application. Summary of the Invention
[0007] To overcome the problems of existing technologies, such as the ease with which the standard PSO algorithm can get trapped in local extrema and poor adaptability when applied to high-dimensional, nonlinear multispectral inversion, the difficulty in adapting to different stages of the optimization process when facing the complex search space of high-dimensional, nonlinear multispectral inversion, the lack of adaptive adjustment mechanism, and low robustness, this invention proposes an improved quantum particle swarm multispectral radiation thermometry method and system.
[0008] To solve the above-mentioned technical problems, the present invention is achieved through the following technical solution: Option 1: This invention proposes an improved multispectral radiation thermometry method for quantum particle swarm optimization, the method comprising the following steps: Step 1: Use a blackbody furnace to calibrate the multi-wavelength device and obtain the response-temperature curve. Align the multi-wavelength device with the target in the heating furnace. The optical signal is transmitted through an optical fiber to a multi-spectral beam modulation box. After photoelectric conversion, the multi-channel voltage signal is obtained by the acquisition card. Step 2: Based on the multi-channel voltage signal obtained in Step 1, construct the objective function for multispectral radiation thermometry; Step 3: Generate initial particle positions using Tent chaotic mapping, and generate the resulting chaotic sequence. Mapping to the solution space range of multispectral inversion This generates the initial particle swarm. Step 4: Calculate the initial population particle fitness and update the initial individual optimum and global optimum, and calculate the average optimal particle position; Step 5: Update the optimal particle position using the Levy flight mechanism; Step 6: Repeat steps 4 and 5 to calculate the objective function value of multispectral radiometric thermometry at the updated location, update the initial individual optimum and global optimum, and output the global optimum if the error threshold or maximum number of iterations is met. ; The method for constructing the objective function for multispectral radiometric thermometry in step 2 is as follows:
[0009] in, The average temperature of each channel. For spectral channel indexing, This represents the total number of spectral channels. To exclude the first Index of other channels besides the one channel, For the first Temperature values obtained from channel inversion For the first Temperature values obtained from channel inversion The weighting coefficient for the penalty term. This is a penalty for exceeding the emission rate range.
[0010] Furthermore, a preferred embodiment is provided, wherein the method for generating the initial particle position using Tent chaotic mapping in step 3 is as follows:
[0011] in, For the first Chaotic variables generated in the next iteration.
[0012] Furthermore, a preferred embodiment is provided, wherein the method for calculating the average optimal particle position in step 4 is as follows:
[0013] in, It is the first The individual best position in the history of each particle, where N is the population size.
[0014] Furthermore, a preferred embodiment is provided, wherein the method for updating the optimal particle position in step 5 using the Lévy flight mechanism is as follows: Step 4.1: Calculate the local attractor ,
[0015] in, It is a random number. It is the globally optimal position; Step 4.2: Utilize the adaptive contraction-expansion coefficient The particle position is updated using a standard normally distributed random number. Step 4.3: Use the decreasing method with the number of iterations to adjust the nonlinear dynamic coefficients. Adjustments are made to cross local extreme points.
[0016] Furthermore, a preferred embodiment is provided in which an adaptive contraction-expansion coefficient is used in step 4,2. The method for updating particle positions using standard normally distributed random numbers is as follows:
[0017] in, , As a local attractor, it is optimal for individual entities. and global optimal Randomly weighted generation, These are Lévy random numbers, generated using the Mantegna algorithm.
[0018] in, The contraction-expansion coefficient, The stability index of the Lévy distribution. For random numbers that follow a uniform distribution, For the particle in the first The position at the next iteration.
[0019] Furthermore, a preferred embodiment is provided in which step 4.3 employs a decreasing method with the number of iterations to adjust the nonlinear dynamic coefficients. The adjustment method is as follows:
[0020] in, The initial contraction-expansion coefficient, To terminate the contraction-expansion coefficient, This represents the current iteration number. This represents the maximum number of iterations.
[0021] Option 2: An improved multispectral radiation thermometry system for quantum particle swarm optimization, the system comprising: The input module is used to calibrate the multi-wavelength device using a blackbody furnace, obtain the response-temperature curve, align the multi-wavelength device with the target to be tested in the heating furnace, and transmit the optical signal through optical fiber to the multi-spectral beam modulation box. After photoelectric conversion, the acquisition card obtains the multi-channel voltage signal. The objective function construction module is used to construct a multispectral radiation thermometry objective function based on the multichannel voltage signals acquired by the data acquisition module. The initialization module is used to generate initial particle positions using the Tent chaotic map and to generate the chaotic sequence. Mapping to the solution space range of multispectral inversion This generates the initial particle swarm. The average optimal particle position calculation module is used to calculate the initial population particle fitness and update the initial individual optimal and global optimal, and to calculate the average optimal particle position. The optimal particle position update module is used to update the optimal particle position through the Levy flight mechanism; The output module repeats the calculation of the average optimal particle position and the update of the optimal particle position module. It calculates the multispectral radiometric thermometry objective function value of the updated position, updates the initial individual optimum and the global optimum, and outputs the global optimum if the error threshold or the maximum number of iterations is met. ; The method for constructing the objective function for multispectral radiometric thermometry in the objective function construction module is as follows:
[0022] in, The average temperature of each channel. For spectral channel indexing, This represents the total number of spectral channels. To exclude the first Index of other channels besides the one channel, For the first Temperature values obtained from channel inversion For the first Temperature values obtained from channel inversion The weighting coefficient for the penalty term. This is a penalty for exceeding the emission rate range.
[0023] Option 3: A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method described in Option 1.
[0024] Option 4: A computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor runs the computer program stored in the memory, the processor executes the method described in Option 1.
[0025] The advantages of this invention are: This invention discloses an improved multispectral radiometric temperature measurement method and system based on quantum particle swarm optimization. It introduces the Tent chaotic mapping within the quantum particle swarm optimization framework to optimize the initial particle positions and further incorporates the Lévy flight mechanism, utilizing its long-tailed super-diffusion properties to enhance global exploration capabilities. This enables rapid and accurate inversion of true temperature and spectral emissivity. The improved multispectral radiometric temperature measurement method and system of this invention features a robust improved optimization algorithm with stronger search capabilities and higher convergence accuracy. It holds significant research and engineering value for improving the inversion performance of multispectral radiometric temperature measurement technology under complex conditions and promoting its wider application.
[0026] This invention is also applicable to fields such as retrieving the true temperature and spectral emissivity of a target from multi-channel spectral radiation signals. Attached Figure Description
[0027] Figure 1 This is a schematic flowchart of an improved quantum particle swarm multispectral radiation thermometry method as described in Embodiment 1.
[0028] Figure 2 This is a schematic diagram of the device principle for an improved quantum particle swarm multispectral radiation thermometry method as described in Embodiment 1. Detailed Implementation
[0029] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them.
[0030] Implementation Method 1, see [link] Figures 1 to 2 This embodiment describes a multispectral radiation thermometry method based on an improved quantum particle swarm optimization technique, specifically including the following steps: Step 1: Construct a multispectral radiation thermometry hardware system and acquire data. This system includes a calibration module, an acquisition module, and a processing module. A blackbody furnace is used to calibrate the multi-wavelength device, obtaining the response-temperature curve. The multi-wavelength device is then aligned with the target in the heating furnace. The optical signal is transmitted via optical fiber to a multispectral beam modulation box, where it is converted into multi-channel voltage signals by a data acquisition card.
[0031] Step 2: Construct the objective function for multispectral radiometric thermometry According to Planck's radiation law, for those with Multispectral radiation pyrometer with multiple wavelength channels Spectral channel output signal It is represented by equation (1).
[0032]
[0033] In equation (1), This is an instrument constant that depends only on wavelength. For temperature The target spectral emissivity, This is the second radiation constant. When it satisfies... When the approximation is made using Wien's law, the formula is shown in equation (2).
[0034]
[0035] Brightness temperature is an equivalent temperature parameter characterizing radiation intensity. It is defined as follows: at a specific wavelength, if the radiation intensity of an actual object is equal to that of a blackbody, then the temperature of the blackbody is considered the brightness temperature of the object at that wavelength. The relationship between the actual temperature and the brightness temperature can be obtained through theoretical derivation, as shown in equation (3).
[0036]
[0037] in, wavelength The brightness temperature is considered. If the temperature in a certain channel deviates significantly from the true value, it will affect the average temperature of the channel, thus adversely affecting subsequent calculations. Furthermore, the temperature of each channel should not affect the average temperature of other channels. Ideally, the temperatures of any two channels should be equal. Based on this, a comprehensive objective function is established. The aim is to minimize the variance of the inversion temperature for each channel and the residual between the measured and theoretical values.
[0038]
[0039] in, The average temperature of each channel. The penalty term for emissivity exceeding the range is defined by a fixed physical characteristic range for emissivity. In the actual algorithm parameter settings, an appropriate penalty range can be selected within this range according to the actual target characteristics.
[0040] In summary, the problem of multispectral radiation thermometry has been transformed into a constrained optimization problem, which can be solved using constrained optimization theory.
[0041] Step 3: Population Initialization Based on Tent Chaotic Mapping To overcome the uneven population distribution and blind search caused by traditional pseudo-random initialization, this invention uses the Tent chaotic map to generate initial particle positions. The Tent map has good ergodicity and uniformity, which allows the initial particles to fully cover the feasible solution space, thereby improving the efficiency of early global search and the stability of the algorithm.
[0042]
[0043] The generated chaotic sequence Mapping to the solution space range of multispectral inversion This generates an initial particle swarm. The Tent map has good ergodicity and uniformity, ensuring that the initial particles fully cover the solution space. Population initialization. Each particle directly represents a set of emission rate vectors. .
[0044] Step 4: Calculate the average optimal position To enhance inter-particle cooperation, an average optimal position is introduced. As the center of the quantum potential field.
[0045]
[0046] in, It is the first The individual best position in the history of each particle, where N is the population size.
[0047] Step 5: Quantum State Update Based on Lévy Flight During the position update phase, the Lévy flight mechanism is introduced. The Lévy distribution has a significant "long tail" characteristic, allowing particles to perform fine-grained searches with small steps while having a certain probability of making long-distance jumps.
[0048] First, calculate the local attractor. .
[0049]
[0050] in, It is a random number. It is the globally optimal position. Then, the adaptive contraction-expansion coefficient is used. The particle position is updated using random numbers from a standard normal distribution (Gaussian distribution). The particle position update formula is as follows:
[0051] in, , As a local attractor, it is optimal for individual entities. and global optimal Generated by random weighting. These are Lévy random numbers, generated using the Mantegna algorithm.
[0052]
[0053] The contraction-expansion coefficient is the sole core parameter controlling the convergence speed of the algorithm. To balance global search and local exploitation, this invention designs a nonlinear dynamic coefficient. The dynamic adjustment strategy decreases with the number of iterations.
[0054]
[0055] in, The initial contraction-expansion coefficient, To terminate the contraction-expansion coefficient, This represents the current iteration number. This represents the maximum number of iterations.
[0056] Compared to linear decreasing, this strategy maintains high quantum activity for a longer period in the early stages of iteration, preventing premature convergence; and converges quickly in the later stages, improving accuracy. Combined with the long-tailed jumping capability of Lévy flight, the algorithm can easily traverse local extrema, making it particularly suitable for solving complex multi-peak surface problems caused by temperature-emissivity coupling in multispectral inversion, significantly improving inversion accuracy.
[0057] Step Six: Iteration and Termination Repeat steps four and five to calculate the fitness value (i.e., the objective function) for the new location. The value of . If the termination condition (error threshold or maximum number of iterations) is met, output the global optimal solution. .
[0058] Example 1: This embodiment provides a specific device for the improved quantum particle swarm multispectral radiation thermometry method, see details below. Figure 2 As shown, the device includes a system calibration module: using an optical probe aligned with a blackbody furnace to perform system calibration of the multi-wavelength device.
[0059] Data acquisition module: The optical probe in the multi-wavelength device is aimed at the target to be tested in the heating furnace. The optical signal is transmitted to the beam splitter through the optical fiber. After photoelectric conversion, the acquisition card acquires the multi-channel voltage signal and obtains the response-temperature curve.
[0060] Data Processing Module: The host computer software loads the acquired data and calls the aforementioned "Improved Quantum Particle Swarm Optimization Algorithm" module. Input: Optimization of the measured voltage parameters for each channel. The algorithm performs simulated evolution calculations in the solution space, when the objective function... Output the result when convergence or when the maximum number of iterations is reached; this is the true temperature obtained through inversion. and emissivity .
[0061] Those skilled in the art will understand that the above description is merely a preferred embodiment of the present invention, and the features described in the various embodiments and / or claims of this disclosure can be combined or combined in various ways, even if such combinations or combinations are not explicitly described in this disclosure. This is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
[0062] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention. Clearly, those skilled in the art can make various alterations and modifications to the invention without departing from its spirit and scope. Thus, if these modifications and modifications of the invention fall within the scope of the claims and their equivalents, the invention is also intended to include these modifications and modifications.
Claims
1. An improved multispectral radiation thermometry method for quantum particle swarm optimization, characterized in that, The method includes the following steps: Step 1: Use a blackbody furnace to calibrate the multi-wavelength device and obtain the response-temperature curve. Align the multi-wavelength device with the target in the heating furnace. The optical signal is transmitted through an optical fiber to a multi-spectral beam modulation box. After photoelectric conversion, the multi-channel voltage signal is obtained by the acquisition card. Step 2: Based on the multi-channel voltage signal obtained in Step 1, construct the objective function for multispectral radiation thermometry; Step 3: Generate initial particle positions using Tent chaotic mapping, and generate the resulting chaotic sequence. Mapping to the solution space range of multispectral inversion This generates the initial particle swarm. Step 4: Calculate the initial population particle fitness and update the initial individual optimum and global optimum, and calculate the average optimal particle position; Step 5: Update the optimal particle position using the Levy flight mechanism; Step 6: Repeat steps 4 and 5 to calculate the objective function value of multispectral radiometric thermometry at the updated location, update the initial individual optimum and global optimum, and output the global optimum if the error threshold or maximum number of iterations is met. ; The method for constructing the objective function for multispectral radiometric thermometry in step 2 is as follows: in, The average temperature of each channel. For spectral channel indexing, This represents the total number of spectral channels. To exclude the first Index of other channels besides the one channel, For the first Temperature values obtained from channel inversion For the first Temperature values obtained from channel inversion The weighting coefficient for the penalty term. This is a penalty for exceeding the emission rate range.
2. The improved quantum particle swarm optimization method for multispectral radiation thermometry according to claim 1, characterized in that, The method for generating the initial particle positions using the Tent chaotic mapping in step 3 is as follows: in, For the first Chaotic variables generated in the next iteration.
3. The improved quantum particle swarm optimization method for multispectral radiation thermometry according to claim 1, characterized in that, The method for calculating the average optimal particle position in step 4 is as follows: in, It is the first The individual best position in the history of each particle, where N is the population size.
4. The improved quantum particle swarm optimization method for multispectral radiation thermometry according to claim 1, characterized in that, The method for updating the optimal particle position using the Lévy flight mechanism in step 5 is as follows: Step 4.1: Calculate the local attractor , in, It is a random number. It is the globally optimal position; Step 4.2: Utilize the adaptive contraction-expansion coefficient The particle position is updated using a standard normally distributed random number. Step 4.3: Use the decreasing method with the number of iterations to adjust the nonlinear dynamic coefficients. Adjustments are made to cross local extreme points.
5. The improved quantum particle swarm optimization method for multispectral radiation thermometry according to claim 4, characterized in that, In step 4.2, the adaptive contraction-expansion coefficient is used. The method for updating particle positions using standard normally distributed random numbers is as follows: in, , As a local attractor, it is optimal for individual entities. and global optimal Randomly weighted generation, These are Lévy random numbers, generated using the Mantegna algorithm. in, The contraction-expansion coefficient, The stability index of the Lévy distribution. For random numbers that follow a uniform distribution, For the particle in the first The position at the next iteration.
6. The improved quantum particle swarm optimization method for multispectral radiation thermometry according to claim 4, characterized in that, Step 4.3 employs a decreasing method with the number of iterations to adjust the nonlinear dynamic coefficients. The adjustment method is as follows: in, The initial contraction-expansion coefficient, To terminate the contraction-expansion coefficient, This represents the current iteration number. This represents the maximum number of iterations.
7. An improved multispectral radiation thermometry system for quantum particle swarm optimization, characterized in that, The system includes: The input module is used to calibrate the multi-wavelength device using a blackbody furnace, obtain the response-temperature curve, align the multi-wavelength device with the target to be tested in the heating furnace, and transmit the optical signal through optical fiber to the multi-spectral beam modulation box. After photoelectric conversion, the acquisition card obtains the multi-channel voltage signal. The objective function construction module is used to construct a multispectral radiation thermometry objective function based on the multichannel voltage signals acquired by the data acquisition module. The initialization module is used to generate initial particle positions using the Tent chaotic map and to generate the chaotic sequence. Mapping to the solution space range of multispectral inversion This generates the initial particle swarm. The average optimal particle position calculation module is used to calculate the initial population particle fitness and update the initial individual optimal and global optimal, and to calculate the average optimal particle position. The optimal particle position update module is used to update the optimal particle position through the Levy flight mechanism; The output module repeats the calculation of the average optimal particle position and the update of the optimal particle position module. It calculates the multispectral radiometric thermometry objective function value of the updated position, updates the initial individual optimum and the global optimum, and outputs the global optimum if the error threshold or the maximum number of iterations is met. ; The method for constructing the objective function for multispectral radiometric thermometry in the objective function construction module is as follows: in, The average temperature of each channel. For spectral channel indexing, This represents the total number of spectral channels. To exclude the first Index of other channels besides the one channel, For the first Temperature values obtained from channel inversion For the first Temperature values obtained from channel inversion The weighting coefficient for the penalty term. This is a penalty for exceeding the emission rate range.
8. A computer storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method described in any one of claims 1-6.
9. A computer device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the program to implement the method of any one of claims 1-6.