Fast convergence electron accelerator output frequency locking method
By determining the resonant frequency range in an electron linear accelerator and using arithmetic sequence and bubble sort methods, the frequency value closest to the resonant frequency can be quickly locked, solving the problems of low search accuracy and long search time in the prior art, and achieving efficient frequency locking.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-31
- Publication Date
- 2026-04-07
AI Technical Summary
In the existing technology, the resonant frequency search of electron linear accelerators is not accurate and takes a long time, making it difficult to quickly lock onto the suboptimal input frequency within the allowable deviation range.
By determining the range of resonant frequency variation, calculating an arithmetic sequence and performing multiple bubble sorts, the frequency search range is gradually narrowed down. Combined with reflection power acquisition and sorting, the frequency value closest to the resonant frequency is quickly locked.
It achieves accurate frequency locking in a very short time, improving the accuracy and efficiency of frequency search and reducing the time cost of frequency locking.
Smart Images
Figure CN121815531A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of space safety technology, specifically relating to a fast-converging electron accelerator output frequency locking method. Background Technology
[0002] Electron linear accelerators have important applications in various fields such as medical and health care, industrial non-destructive testing, food irradiation, radiation imaging, and space particle weapons. Electron linear accelerators have their own resonant frequency. When an input signal is given to the accelerator, reflected power is obtained. The reflected power reaches its minimum and the output power reaches its maximum when the frequency of the input signal equals the resonant frequency; this is the optimal operating condition. The further the frequency deviates from the resonant frequency, the greater the reflected power. During accelerator operation, parameters such as temperature change dynamically, and the resonant frequency changes accordingly. Therefore, during operation, the input frequency needs to be adjusted in real time according to the reflected power to keep it equal to the resonant frequency. On the other hand, perfectly equaling the resonant frequency is the ideal optimal input. Within the entire frequency input range, the area near the resonant frequency can be considered a suboptimal input within the allowable deviation range of the optimal input. The time required for the input frequency to re-lock onto the resonant frequency after reaching the optimal operating condition is also extremely important; the shorter the time, the better. Currently, there is an urgent need both domestically and internationally for methods to quickly bring the input frequency to a suboptimal range within the allowable deviation range. Improving the accuracy of the frequency generator in searching for the resonant frequency of the electron linear accelerator, and significantly increasing the search speed, are directions that urgently need research. Summary of the Invention
[0003] The purpose of this invention is to provide a fast-converging electron accelerator output frequency locking method to overcome the problems of low accuracy of the resonant frequency searched by existing methods and long search time.
[0004] In a first aspect, the present invention provides a fast-converging electron accelerator output frequency locking method, comprising the following steps: Step 1: Determine the range of resonant frequency variation based on the current operating temperature and humidity of the electron linear accelerator, and denot it as the frequency range; Step 2: Using the minimum frequency value obtained in Step 1 as the first term and the maximum frequency value as the last term, calculate an arithmetic sequence of 1000 frequency values. Since one of these 1000 frequency values is closest to the resonant frequency of the electron linear accelerator, find the frequency value closest to the resonant frequency from these 1000 values, or find the frequency value closest to the resonant frequency within the allowable deviation range; this is frequency locking. Divide the 1000 frequency values in the frequency range into 10 frequency segments, and select the middle frequency value from the 100 frequency values in each segment, selecting a total of 10 frequency values. The frequency generator sequentially generates sinusoidal signals of the 10 selected frequency values and provides them to the electron linear accelerator, while simultaneously collecting the corresponding reflected power as the first 10 reflected power samples. Step 3: Perform bubble sort on the 10 reflected power values collected in the first step, select the frequency band corresponding to the smallest reflected power, and further divide the 100 frequency values in this frequency band into 10 smaller frequency bands. Select the middle frequency value from the 10 frequency values in each smaller frequency band, and select a total of 10 frequency values. The frequency generator sequentially generates sinusoidal signals of these 10 frequency values and provides them to the electron linear accelerator, while simultaneously collecting the corresponding reflected power values, which are then used as the 10 reflected power values collected in the second step. Step 4: Perform bubble sort on the 10 reflected power values collected in the second step, select the smaller frequency band corresponding to the minimum reflected power. This frequency band currently has only 10 frequency values, so it is no longer divided equally. The frequency generator sequentially generates sinusoidal signals of these 10 frequency values and provides them to the electron linear accelerator. At the same time, the reflected power is collected as the 10 reflected power values collected in the third step. Step 5: Perform bubble sorting on the 10 reflected power values collected in the third sampling, select the frequency value corresponding to the minimum reflected power, and use this frequency value as the lock frequency. Step 6: Continuously input the locked frequency signal obtained in Step 5 into the electron linear accelerator, and monitor the reflected power. As the resonant frequency of the electron linear accelerator gradually changes with environmental factors, the reflected power corresponding to the locked frequency will become larger and larger until it exceeds the preset acceptable threshold. At this time, it indicates that the new resonant frequency deviates significantly from the locked frequency. At this time, return to Step 1.
[0005] In a second aspect, the present invention provides an electronic device, the electronic device comprising: Memory, used to store executable instructions; A processor, when executing executable instructions or computer programs stored in the memory, implements the method provided in the first aspect of the present invention.
[0006] Thirdly, the present invention provides a computer-readable storage medium storing executable instructions or a computer program, wherein the executable instructions, when executed by a processor, implement the method provided in the first aspect of the present invention.
[0007] This invention, through extensive experimentation, verifies the characteristics of the reflected power curve within the resonant frequency range described in the operating manual of the electron linear accelerator. Based on the curve characteristics and the acceptable deviation requirements of the resonant frequency, it utilizes the principle of monotonicity of quadratic functions, the principle of slope of similar triangles in abstract geometric construction, and topological recursive analysis. It refines the implementation efficiency of FPGA programming and the ease of upgrading and maintaining the method, resulting in a specific frequency locking method for this invention. Laboratory-scale verification of the algorithm has been conducted, demonstrating that the input frequency can be relocked within a very short time after a resonant frequency deviation, bringing it close to the resonant frequency within the allowable deviation range. This shows that the method of this invention can effectively improve accuracy and search efficiency. This invention has strong engineering feasibility, excellent performance, and a broad application market. Attached Figure Description
[0008] Figure 1 This is a flowchart of the fast-convergence electron accelerator output frequency locking method of the present invention; Figure 2 In this embodiment of the invention, 1000 full-frequency points are used to collect reflected power. Figure 3 This is a schematic diagram of the critical deviation situation. Figure 4 The digital circuit used in the experiment was used to verify the results. Detailed Implementation
[0009] like Figure 1 As shown, the fast-convergence electron accelerator output frequency locking method provided by this invention includes the following steps: Step 1: Based on the current operating temperature and humidity of the electron linear accelerator, determine the range of resonant frequency variation, denoted as the frequency range. The frequency generator will then calculate and search for the locked frequency within this range.
[0010] Step 2: Using the minimum frequency value obtained in Step 1 as the first term and the maximum frequency value as the last term, calculate an arithmetic sequence of 1000 frequency values. Since one of these 1000 frequency values is closest to the resonant frequency of the electron linear accelerator, find the frequency value closest to the resonant frequency from these 1000 values, or find the frequency value closest to the resonant frequency within the allowable deviation range; this is frequency locking. Divide the 1000 frequency values in the frequency range into 10 frequency segments. Select the middle frequency value from the 100 frequency values in each segment, selecting a total of 10 frequency values. The frequency generator sequentially generates sinusoidal signals of the 10 selected frequency values and provides them to the electron linear accelerator, while simultaneously collecting the corresponding reflected power as the first 10 reflected power samples. Step 3: Perform bubble sort on the 10 reflected power values collected in the first step, select the frequency band corresponding to the smallest reflected power, and further divide the 100 frequency values in this frequency band into 10 smaller frequency bands. Select the middle frequency value from the 10 frequency values in each smaller frequency band, and select a total of 10 frequency values. The frequency generator sequentially generates sinusoidal signals of these 10 frequency values and provides them to the electron linear accelerator, while simultaneously collecting the corresponding reflected power values, which are then used as the 10 reflected power values collected in the second step. Step 4: Perform bubble sort on the 10 reflected power values collected in the second step, select the smaller frequency band corresponding to the minimum reflected power. This frequency band currently has only 10 frequency values, so it is no longer divided equally. The frequency generator sequentially generates sinusoidal signals of these 10 frequency values and provides them to the electron linear accelerator. At the same time, the reflected power is collected as the 10 reflected power values collected in the third step. Step 5: Perform bubble sorting on the 10 reflected power values collected in the third sampling, select the frequency value corresponding to the minimum reflected power, and use this frequency value as the lock frequency.
[0011] Step 6: Continuously input the signal of the locked frequency obtained in step 5 into the electron linear accelerator, and monitor the reflected power. As the resonant frequency of the electron linear accelerator gradually changes with environmental factors (temperature, humidity, etc.), the reflected power corresponding to the locked frequency will become larger and larger until it exceeds the preset acceptable threshold. At this time, it indicates that the new resonant frequency deviates significantly from the locked frequency. At this time, return to step 1.
[0012] The basis for setting several key parameters and the resulting technical effects in the method of the present invention described above are as follows: (1) The maximum range of resonant frequency variation in step 1: The resonant frequency of an electron linear accelerator dynamically changes with parameters such as temperature and humidity during operation. It is essential to carefully refer to its instruction manual, prepare instruments such as thermometers and hygrometers, and calculate the possible maximum and minimum resonant frequencies to obtain a reasonable resonant frequency range, thereby achieving an accurate lock-on frequency. Figure 2 This represents the measured reflected power of the electron linear accelerator. The horizontal axis represents all 1000 input frequencies, and the vertical axis represents the power amplitude. Point A is the minimum reflected power corresponding to the resonant frequency. The curve to the left of point A shows a monotonically decreasing trend, while the curve to the right shows a monotonically increasing trend. This is a fundamental characteristic of the electron linear accelerator and is the basis of the method of this invention. If the resonant frequency range is incorrect, the curves of the reflected power corresponding to each frequency within the incorrect range will be irregular, causing the method of this invention to fail.
[0013] (2) Input frequency step value and convergence efficiency in steps 2 to 4 The total resonant frequency range is divided into 1000 minimum frequency divisions. To lock onto the final frequency value, the range is divided into 10 equal divisions twice. In steps 2 to 4, only 30 frequency values are input to the electron linear accelerator. Compared to inputting all 1000 frequency values to obtain the reflected power, and using bubble sort to obtain the frequency with the minimum reflected power, the convergence speed is reduced from 1000 times to 10*lg(1000) times. Since each division is into 10 equal divisions, the programming efficiency and accuracy of the frequency generator are greatly optimized. Furthermore, powers of 10 provide an excellent foundation for upgrading or simplifying the method of this invention. To improve accuracy, the range can be divided into 10000 minimum frequency divisions; to reduce accuracy and save costs, the range can be divided into 100 minimum frequency divisions. The principle and operation method of rapid convergence are the same as those of dividing the range into 1000 minimum frequency divisions in this invention.
[0014] (3) The deviation of the locking frequency obtained in steps 1 to 5 is small and the accuracy is high. In the method of this invention, the deviation between the frequency value locked in each search of 1000 frequency values and the frequency value closest to the resonant frequency among these 1000 frequencies does not exceed 5 frequency points, that is, the deviation is less than or equal to 0.5%. The deviation between the frequency value closest to the resonant frequency among the 1000 frequencies and the actual resonant frequency is determined by the minimum division value. The method of this invention divides the resonant frequency range into 1000 equal frequency points, so the deviation is 0.1%.
[0015] The explanation and proof for 0.5%, i.e., a deviation of no more than 5 frequency points, are as follows: The core of this invention lies in the ability of each decimation bubble sort to identify the frequency band containing point A. This ensures that the third bubble sort can accurately locate point A with zero deviation. If the first or second sort incorrectly selects a frequency band far from point A (e.g., ...), the problem is mitigated. Figure 2If the horizontal axis is in the frequency range of 600 to 700, then the final locked point deviates from point A by more than 5. This situation can be ruled out because there is a clear monotonic trend around point A. Only critical cases need to be considered, such as... Figure 3 The input signal frequencies selected for the 500-600 and 600-700 frequency bands are 550 and 650 MHz, respectively. The reflected power at these two points is extremely close. If the 500-600 MHz band were mistakenly selected, the monotonic trend indicates that the point chosen in the third bubble sort is the rightmost point in this band. The maximum deviation of this point from point A is 5. The horizontal axis distance from 550 to point A is 55, and the distance from 650 to point A is 45. The slope ratio 55:45 = 11:9 ensures that the reflected power at points 550 and 650 MHz is exactly equal. The maximum difference in slope between the left and right sides of point A not exceeding 11:9 is a characteristic of the electron accelerator's resonance curve. If the deviation is greater than 5, the horizontal axis difference will be greater than 55:45, resulting in a greater reflected power at 550 MHz than at 650 MHz, thus preventing the 500-600 MHz band from being mistakenly selected. Therefore, the first bubble sort causes a maximum deviation of 5 frequency points. Similarly, due to the slope difference, the second maximum deviation will only be 1 frequency point (horizontal axis 11:9), which will not be elaborated here.
[0016] To verify the feasibility and effectiveness of this invention, the following laboratory verification experiments are presented. The digital circuit implementation used in the experiments is as follows: Figure 4 As shown, the left half is Figure 1 The actual implementation of the frequency generator. The method process of this invention is implemented using VHDL statements on an FPGA.
[0017] The resonant frequency of the electron linear accelerator varies with temperature, with a minimum of 9.29 GHz and a maximum of 9.31 GHz, which can be obtained from the equipment manual. A series of 1000 frequency values, from 9.29 GHz to 9.31 GHz, are formed, numbered sequentially from 1 to 1000. For example, frequency number 1 corresponds to 9.29 GHz. The theoretical resonant frequency at a given temperature is obtained from the equipment manual; among the 1000 discrete frequency values, one is closest to the theoretical resonant frequency. The temperature of the chamber was gradually adjusted from 0℃ to 40℃, and the experimental records are shown in Table 1.
[0018] Table 1
[0019] As can be seen from the results in Table 1, the maximum deviation rate of the locked frequency of the method of the present invention does not exceed 0.5%, which is more than 10 percentage points higher than other methods in the industry. This demonstrates that the resonant frequency obtained by the fast-convergence electron accelerator output frequency locking method of the present invention has high accuracy.
[0020] In addition, the method of the present invention has a fast convergence speed, and the frequency values of 1000 discrete points can be converged in only 30 times, which is more than 25 percentage points ahead of other methods in the industry; the parameters and architecture adopt a logarithmic form with base 10, which greatly improves the convenience of use, maintenance and upgrade.
Claims
1. A fast-converging electron accelerator output frequency locking method, characterized in that, Includes the following steps: Step 1: Determine the range of resonant frequency variation based on the current operating temperature and humidity of the electron linear accelerator, and denot it as the frequency range; Step 2: Using the minimum frequency value obtained in Step 1 as the first term and the maximum frequency value as the last term, calculate an arithmetic sequence of 1000 frequency values. Since one of these 1000 frequency values is closest to the resonant frequency of the electron linear accelerator, find the frequency value closest to the resonant frequency from these 1000 values, or find the frequency value closest to the resonant frequency within the allowable deviation range; this is frequency locking. Divide the 1000 frequency values in the frequency range into 10 frequency segments, and select the middle frequency value from the 100 frequency values in each segment, selecting a total of 10 frequency values. The frequency generator sequentially generates sinusoidal signals of the 10 selected frequency values and provides them to the electron linear accelerator, while simultaneously collecting the corresponding reflected power as the first 10 reflected power samples. Step 3: Perform bubble sort on the 10 reflected power values collected in the first step, select the frequency band corresponding to the smallest reflected power, and further divide the 100 frequency values in this frequency band into 10 smaller frequency bands. Select the middle frequency value from the 10 frequency values in each smaller frequency band, and select a total of 10 frequency values. The frequency generator sequentially generates sinusoidal signals of these 10 frequency values and provides them to the electron linear accelerator, while simultaneously collecting the corresponding reflected power values, which are used as the 10 reflected power values collected in the second step. Step 4: Perform bubble sort on the 10 reflected power values collected in the second step, select the smaller frequency band corresponding to the minimum reflected power. This frequency band currently has only 10 frequency values, so it is no longer divided equally. The frequency generator sequentially generates sinusoidal signals of these 10 frequency values and provides them to the electron linear accelerator. At the same time, the reflected power is collected as the 10 reflected power values collected in the third step. Step 5: Perform bubble sorting on the 10 reflected power values collected in the third sampling, select the frequency value corresponding to the minimum reflected power, and use this frequency value as the lock frequency. Step 6: Continuously input the locked frequency signal obtained in Step 5 into the electron linear accelerator, and monitor the reflected power. As the resonant frequency of the electron linear accelerator gradually changes with environmental factors, the reflected power corresponding to the locked frequency will become larger and larger until it exceeds the preset acceptable threshold. At this time, it indicates that the new resonant frequency deviates significantly from the locked frequency. At this time, return to Step 1.
2. An electronic device, characterized in that, The electronic device includes: Memory, used to store executable instructions; A processor, when executing executable instructions or computer programs stored in the memory, implements the method as described in claim 1.
3. A computer-readable storage medium storing executable instructions or a computer program, characterized in that, When the executable instructions are executed by the processor, they implement the method as described in claim 1.