A system and method for optimizing performance of an accelerator superconducting main magnet power supply

Abstract

The application discloses an accelerator superconducting main magnet power supply performance optimization system and method, the system comprises a 232 communication conversion network communication unit, a circuit modeling unit based on damping ratio verification, a parameter setting unit based on damping ratio and open loop gain and steady state error; the method comprises the following steps: converting the communication mode of the main magnet power supply loop which is close and susceptible to electromagnetic interference into the main magnet network communication mode which is long-distance and anti-interference, and obtaining high-precision current information of the remote main magnet power supply loop; circuit modeling based on high-precision current information and damping ratio verification; parameter setting based on damping ratio and open loop gain and steady state error; the application avoids the problems of too short 232 serial communication distance, susceptibility to electromagnetic, radiation and other environmental interference; the modeling method reduces the iteration number of parameter setting; by improving the system open loop gain and increasing the damping ratio, the problem that the output current stability is difficult to improve due to the large superconducting magnet load inductance and time constant is solved.

Description

Technical Field

[0001] This invention belongs to the field of accelerator superconducting main magnet power supply control in proton therapy, and more specifically, relates to an accelerator superconducting main magnet power supply performance optimization system and method. Background Technology

[0002] The main magnet is the core structure of a cyclotron, used to generate a powerful, stable, and uniform magnetic field to control the motion of charged particles within the accelerator. For medium- to high-energy isochronous superconducting cyclotrons, particles will undergo nearly a thousand revolutions, which places extremely stringent requirements on the accumulation of slip phases and also demands a high stability of 10 ppm from the superconducting magnet power supply.

[0003] The stability of the output current of a superconducting magnet power supply is highly sensitive to environmental factors. Over several years of operation, slight variations in the superconducting coil load, power supply line resistance, and changes in environmental factors such as temperature and humidity can decrease the stability of the output current, significantly impacting beam extraction efficiency. Therefore, regular monitoring and performance improvements of the power supply's output current are necessary. However, improving the output performance of a superconducting magnet power supply presents the following challenges:

[0004] (1) The accelerator facility is large, and the main control room and power supply equipment are far apart. At the same time, the electromagnetic and radiation environments are complex. This makes it difficult to obtain high-precision current monitoring data while controlling the operation of the accelerator.

[0005] (2) Control parameters cannot be adjusted online. Superconducting magnets have extremely high magnetic energy storage, reaching hundreds of MJ. If the control parameters are not adjusted properly, the output current will oscillate significantly, and a large amount of energy will be rapidly transferred between the power supply and the load, posing a risk of damaging the power supply and the superconducting load. Therefore, the parameters can only be adjusted after the load has been completely demagnetized and then re-excited, a process that will last for several hours. If the parameters are adjusted based on experience, it often requires many iterations, greatly reducing efficiency and resulting in wasted effort.

[0006] (3) The inductance of a superconducting magnet load can reach hundreds of ohms, while the resistance is very small. Therefore, the time constant of the load is very large, which is a large time delay for the control system. This makes it difficult for the control system to respond quickly to disturbances, making it difficult to improve the stability of the output current. Summary of the Invention

[0007] To address the problems existing in the prior art, this invention proposes a system and method for optimizing the power supply performance of the superconducting main magnet in an accelerator. The first objective is to solve the problem that, due to the distance between the accelerator control unit and the superconducting magnet power supply equipment, and the presence of electromagnetic and radiation interference loads, it is difficult to obtain high-precision current monitoring data while the accelerator is operating. The second objective is to solve the problem that the control parameters cannot be adjusted online due to the extremely large energy storage of the superconducting coil load. The third objective is to solve the problem that, due to the large inductance and time constant of the superconducting magnet load, the control system's response to disturbances is lag, making it difficult to improve the stability of the output current.

[0008] To address the problems existing in the prior art, the present invention proposes the following technical solution:

[0009] An accelerator superconducting main magnet power supply performance optimization system includes a 232 communication-to-network communication unit, a circuit modeling unit based on damping ratio verification, and a parameter tuning unit based on damping ratio, open-loop gain, and steady-state error; its features are:

[0010] The 232 communication-to-network communication unit is used to convert the short-range, electromagnetically susceptible communication method into a long-range, interference-resistant network communication method, so as to remotely obtain high-precision main magnet power supply information and send the high-precision current measurement data of the main magnet power supply to the circuit modeling unit based on damping ratio verification.

[0011] The circuit modeling unit based on damping ratio verification is used to simulate and model the power supply circuit of the superconducting magnet based on the high-precision current information and the main magnet power supply circuit information, to obtain a circuit model that can reflect the real system, and to send the circuit model to the parameter tuning unit based on damping ratio, open-loop gain and steady-state error.

[0012] The parameter tuning unit based on damping ratio, open-loop gain, and steady-state error is used to derive tuning parameters that increase the damping ratio, increase the open-loop gain, and decrease the steady-state error based on the transfer function of the circuit model that reflects the real system after obtaining the circuit model. These parameters are then used to control the main magnet power supply circuit, thereby improving the performance of the main magnet power supply in the circuit. The low-precision main magnet power supply information refers to the low-precision main magnet power supply information caused by electromagnetic and radiation interference in the near-field environment. The high-precision main magnet power supply information refers to the high-precision main magnet power supply information formed by using network communication to move away from electromagnetic and radiation interference.

[0013] Furthermore, the 232 communication to network communication unit is sequentially configured with: a 232 to network communication subunit, an equipment-side Ethernet switch subunit, a control room Ethernet switch subunit, a main control room real-time analysis subunit, and a current measurement data subunit. The 232 to network communication subunit includes a DCCT, a high-precision multimeter, and a virtual serial port module. High-precision current measurement values ​​are obtained through the DCCT and high-precision multimeter, and then connected to the virtual serial port module, which in turn connects to the network switch, enabling the high-precision current measurement values ​​to be transmitted via Ethernet. This extends the transmission distance and reduces interference. The main control room real-time analysis subunit transmits the high-precision current data to the main control computer in real time. The main control computer's analysis program then plots the real-time current change graph and analyzes and calculates various performance parameters of the output current.

[0014] Furthermore, the circuit modeling unit based on damping ratio verification is sequentially configured with: a power supply loop modeling subunit, a model damping ratio calculation subunit, a model error control subunit, and a model output subunit. The power supply loop modeling subunit uses simulation software to model the power supply loop between the power source and the superconducting magnet, and outputs the model to the model damping ratio calculation subunit. The model error control subunit compares the calculated model damping ratio with the actual system damping ratio, controlling the error to within 5%, thereby obtaining a circuit model that reflects the real system. The model output subunit sends the model that reflects the real system to the parameter tuning unit based on the damping ratio, open-loop gain, and steady-state error.

[0015] Furthermore, the power supply circuit modeling subunit includes a power controller type acquisition module, a main magnet load parameter acquisition module, and a power supply circuit modeling module. The power supply circuit modeling module uses simulation software to model the power supply circuit containing the power controller type and main magnet load parameters.

[0016] Furthermore, the control model error subunit includes a module for calculating the damping ratio of the actual system, a module for obtaining the damping ratio of the calculated model, a damping ratio error comparison module, and a control model accuracy module. The control model accuracy module determines whether to remodel based on the comparison result of the damping ratio error comparison module. If the error is less than 5%, the established model is considered to reflect the parameters of the real system; otherwise, the power supply circuit modeling subunit will remodel.

[0017] Furthermore, the parameter tuning unit based on damping ratio, open-loop gain, and steady-state error includes a control parameter derivation subunit, a control parameter adjustment subunit, and a control parameter tuning subunit. The control parameter derivation subunit derives the relationship between the control parameters and the model damping ratio, open-loop gain, and steady-state error based on the model's transfer function. The control parameters include PID parameters. The adjustment subunit adjusts the PID parameters according to the derived relationship, appropriately increasing the model damping ratio, appropriately increasing the open-loop gain, and decreasing the steady-state error, thereby completing the parameter tuning. The control parameter tuning subunit performs output current simulation on the tuned parameters and observes the simulation effect. If the performance and system stability meet the set requirements, the tuned parameters are input into the main magnet power supply.

[0018] Furthermore, the appropriate increase in model damping ratio, appropriate increase in open-loop gain, and reduction in steady-state error are specifically as follows: The principle for appropriately increasing the damping ratio is that, after the increase, the response characteristics of the output current in the simulation results should not become overdamped, and the current overshoot should not completely disappear; the principle for appropriately increasing the open-loop gain is that, after the increase, the stability of the model should not be negative when using methods such as the Routh criterion to evaluate it; simultaneously, the oscillation settling time should not increase in the simulation results of the model output current; the principle for reducing the steady-state error is that the steady-state error of the model should be reduced to 0 as much as possible, and if it cannot be reduced to 0, it should be minimized as much as possible, and the overall picture should be considered, ensuring that the damping ratio of the model increases while reducing the steady-state error.

[0019] Furthermore, the performance and system stability meet the set requirements, specifically: based on the difference between the output current stability and the target value in the analyzed power supply performance, if the difference is within 5 times, the output current stability of the model after parameter tuning should be increased by the corresponding factor; if the difference is more than 5 times, the output current stability of the model after parameter tuning should be increased by at least 5 times; at the same time, the output current of the model after parameter tuning should remain stable and should not be out of control.

[0020] A method for optimizing the power supply performance of a superconducting main magnet in an accelerator, characterized by the following steps:

[0021] Step 1: Convert the communication method of the main magnet power supply circuit, which is close-range and susceptible to electromagnetic interference, into a long-range and interference-resistant main magnet network communication method, and obtain high-precision current information of the remote main magnet power supply circuit;

[0022] Step 2: Circuit modeling based on high-precision current information and damping ratio verification: Calculate the model damping ratio and the actual system damping ratio based on the acquired high-precision current data. Compare the two as a criterion for whether the model can reflect the actual circuit system. If the difference is less than 5%, the model is considered to reflect the actual circuit system; otherwise, the model should be remodeled.

[0023] Step 3: Parameter tuning based on damping ratio, open-loop gain and steady-state error: Adjust the control parameters in the obtained circuit model, and judge whether the performance has been improved through simulation analysis, so as to obtain better control parameters.

[0024] Furthermore, adjusting the control parameters in the obtained circuit model specifically involves: appropriately increasing the model damping ratio, appropriately increasing the model open-loop gain, and reducing the model steady-state error, thereby completing the parameter tuning; and performing current simulation on the tuned parameters and observing the simulation effect. If the performance and system stability meet the set requirements, the tuned parameters are then input into the main magnet power supply.

[0025] Advantages and effects of the present invention

[0026] 1. This invention proposes a method for building a remote high-precision current monitoring platform, which avoids the problems of short communication distance and susceptibility to electromagnetic and radiation interference caused by multimeters through the 232 serial port. It enables the acquisition and analysis of high-precision current monitoring data while controlling the operation of the accelerator, laying the foundation for performance improvement.

[0027] 2. The modeling method for the superconducting magnet load power supply circuit proposed in this invention can perform control parameter tuning attempts and effect verification on a model that closely resembles the actual system, reducing the number of iterations for parameter tuning, saving the time of repeated demagnetization and re-excitation, and greatly improving work efficiency.

[0028] 3. This invention proposes an optimization strategy for control parameters. By increasing the open-loop gain of the system and increasing the damping ratio, it solves the problem that the control system's response to disturbances is lagging due to the large load inductance and time constant of the superconducting magnet, which makes it difficult to improve the stability of the output current. Attached Figure Description

[0029] Figure 1 This is a functional block diagram of the accelerator superconducting main magnet power supply performance optimization system of the present invention;

[0030] Figure 2 This is a functional block diagram of the 232 communication-to-network communication unit of the present invention;

[0031] Figure 3 This is a functional block diagram of the modeling unit for the present invention based on high-precision current information and damping ratio verification.

[0032] Figure 4 This is the parameter tuning unit of the present invention based on damping ratio, open-loop gain and steady-state error;

[0033] Figure 5 This is a flowchart of the method for optimizing the power supply performance of the superconducting main magnet in the accelerator according to the present invention.

[0034] Figure 6aBefore optimization, the power supply output current stability was 87.33ppm, far exceeding the required 10ppm schematic diagram;

[0035] Figure 6b Before optimization, the oscillation adjustment time after the power supply output current overshoot was as long as 851s (illustrated diagram).

[0036] Figure 6c To optimize the simulation results of the output current before optimization: overall diagram of the current rise process and steady-state process;

[0037] Figure 6d To optimize the simulation results of the output current before optimization: an enlarged view of the steady-state process;

[0038] Figure 6e The simulation results of the optimized output current are shown in the overall diagram of the current rise process and steady-state process.

[0039] Figure 6e The simulation results of the optimized output current are shown in the overall diagram of the current rise process and steady-state process.

[0040] Figure 6f An enlarged view of the steady-state process of the optimized output current simulation results;

[0041] Figure 6g A comparison chart of the oscillation adjustment time after power supply output current overshoot before and after parameter tuning;

[0042] Figure 6h The diagram shows how the power supply output current stability improved to 8.30ppm after parameter tuning. Detailed Implementation

[0043] I. Innovation of this invention

[0044] The innovation lies in achieving unexpected results by changing the position of the elements and the relative relationships between them.

[0045] The position of the transformed element was changed from close-range measurement to long-range measurement, thereby avoiding the influence of environmental interference around the accelerator on the current and obtaining a "clean" current, which laid the foundation for subsequent modeling.

[0046] The change in the relative relationships between elements is a shift from direct to indirect debugging. Direct debugging faces unavoidable factors. For example, significant oscillations may occur during current adjustment. Since superconducting magnets are highly sensitive to rapid changes in current, large-amplitude current oscillations may cause the superconducting magnet to lose its quench, leading to danger. Another example is that the inductance of a superconducting magnet load can reach hundreds of ohms, while the resistance is very small. Therefore, the time constant of the load is very large, which is a large time delay for the control system. This prevents the control system from responding quickly to disturbances. These are all unavoidable factors associated with direct debugging.

[0047] This invention employs a modeling method. First, debugging is performed within the model, thus avoiding significant oscillations that can occur during on-site current debugging. On-site debugging can lead to substantial output current oscillations due to instability caused by improper parameter tuning. Debugging within the model allows observation of system instability through output current simulation results. If instability is detected, parameters should be modified to obtain a suitable set of control parameters. Applying these suitable parameters to the actual system will prevent current instability. A new problem with indirect debugging is that inaccurate models cannot achieve the same effect as direct debugging. The solution is to use a comparison method based on damping ratio. Although current is debugged within the model, the obtained current is a "clean" current. With a clean current, the actual damping ratio can be calculated. The actual damping ratio is used as a benchmark to measure the modeling effect. If the error between the actual damping ratio and the damping ratio calculated by the simulation software exceeds a certain range, the model is rebuilt, thus ensuring parameter accuracy. Secondly, the parameters are tuned within the model, directly controlling the power supply of the main magnet's power circuit. This eliminates the need for direct power supply adjustments, preventing significant oscillations during the adjustment process. Even if oscillations occur, they are confined to the model; once the parameters are properly adjusted, oscillations will cease, only occurring in cases of overshoot or undershoot.

[0048] II. Design Principles of the Invention

[0049] 1. Design principle of RS-232 communication to network communication

[0050] The limitations of RS-232 communication lie in its effective transmission distance of only 15 meters. The main magnet power supply and the main control room PC are typically located on different floors, often more than 15 meters apart, exceeding the effective communication distance and causing severe data distortion. Furthermore, the accelerator itself is very large, comparable to a large factory building. This massive device involves numerous components and strong magnetic radiation. Even if the distance between the main magnet power supply and the main control room PC is within 15 meters, the transmitted data from the main magnet power supply circuit will inevitably be affected by environmental interference, resulting in non-clean current data and low-precision transmission. Improving RS-232 communication to network communication leverages the fact that network cables can penetrate walls, which act as excellent magnetic shielding layers, thus avoiding electromagnetic and radiation interference during data transmission. Moreover, the effective transmission distance of a network cable is around 100 meters; improving RS-232 communication to network communication extends the data transmission distance. A key aspect of converting RS-232 communication to network communication is the use of a virtual serial port module. The module takes a RS-232 serial communication cable as input and a network communication cable as output. It connects to a local area network (LAN) containing the control room PC, allowing the PC to access the module via the network. Furthermore, the module can create a virtual serial port within its LAN. Therefore, when the control room PC accesses the device, it is essentially connected to this virtual serial port and can receive data from the RS-232 serial communication via the network.

[0051] 2. Design principles for model building based on high-precision current and damping ratio verification

[0052] The high-precision current referred to is a "clean" current, unaffected by electromagnetic interference, radiation, or other environmental disturbances. A clean current accurately reflects the parameters of the main magnet power supply circuit. The first key point of modeling is obtaining a "clean" current; the second key point is ensuring the model is scientific and accurately describes the parameters of the main magnet power supply circuit. This invention employs the following method: calculating the actual damping ratio using the "clean" current, calculating the theoretical damping ratio using simulation software, and then comparing these two. If the error is less than 5%, the model accurately reflects the main magnet power supply circuit. The damping ratio verification method ensures the model's scientific validity. A "clean" current is a prerequisite. The relationship between the two is that modeling is only accurate and scientific based on a "clean" current, but a "clean" current does not guarantee a scientifically sound model. Therefore, a damping ratio verification method is also necessary. Both are indispensable and interdependent.

[0053] 3. Parameter tuning design principle based on damping ratio, open-loop gain, and steady-state error

[0054] ① This invention addresses a problem from three aspects: Through modeling, it solves the problem of "a large time constant and slow response speed of the output current to the control system" during the process of current changing from its current value to the target value in three ways. First, it increases the damping ratio. Increasing the damping ratio increases the resistance during the oscillation process around the target value after the current reaches its vicinity. A larger resistance results in smaller current fluctuations, shortens the oscillation adjustment process, and relatively reduces the time constant during the change of current from the current value to the target value. Second, it increases the open-loop gain. Increasing the open-loop gain increases the system's control quantity. System control quantity * error = control effect. Therefore, with a constant error, a larger system control quantity results in a better effect, and a better effect leads to a relatively smaller time constant during the change of current from the current value to the target value. Third, it reduces steady-state error. During steady-state output, the smaller the error between the target value and the actual current value, the smaller the time constant during the change of current to the target value.

[0055] In summary, increasing the open-loop gain, increasing the damping ratio, and reducing the steady-state error are like solving a problem from three different directions. Increasing the open-loop gain is like increasing the thrust from the positive direction, increasing the damping ratio is like increasing the resistance from the reverse direction when the current fluctuates around the target value, so that the current fluctuation stops quickly, and reducing the steady-state error is like compressing the vertical error from the vertical direction.

[0056] ② Principles for increasing damping ratio, increasing open-loop gain, and reducing steady-state error. 1) The principle for appropriately increasing the damping ratio is: after increasing it, the response characteristics of the output current in the simulation results should not become overdamped, because overdamping will slow down the current response speed to the control system and increase the time constant. In short: there should be some overshoot, but the overshoot control should not completely disappear. 2) The principle for appropriately increasing the open-loop gain is: after increasing it, the stability of the model should be evaluated using methods such as the Routh criterion; the stability should not be negative. At the same time, the oscillation settling time should not increase in the simulation results of the model's output current. 3) The principle for reducing steady-state error is: the steady-state error of the model should be reduced to 0 as much as possible. If it cannot be reduced to 0, it should be minimized as much as possible, and the overall picture should be considered. While reducing the steady-state error, the damping ratio of the model must be increased.

[0057] Based on the above principles, this invention designs a power supply performance optimization system for accelerator superconducting main magnets, such as... Figure 1 , Figure 2 , Figure 3 , Figure 4 As shown, it includes a 232 communication to network communication unit, a circuit modeling unit based on damping ratio verification, and a parameter tuning unit based on damping ratio, open-loop gain, and steady-state error; its features are:

[0058] The 232 communication-to-network communication unit is used to convert the short-range, electromagnetically susceptible communication method into a long-range, interference-resistant network communication method, so as to remotely obtain high-precision main magnet power supply information and send the high-precision current measurement data of the main magnet power supply to the circuit modeling unit based on damping ratio verification.

[0059] The circuit modeling unit based on damping ratio verification is used to simulate and model the power supply circuit of the superconducting magnet based on the high-precision current information and the main magnet power supply circuit information, to obtain a circuit model that can reflect the real system, and to send the circuit model to the parameter tuning unit based on damping ratio, open-loop gain and steady-state error.

[0060] The parameter tuning unit based on damping ratio, open-loop gain, and steady-state error is used to derive tuning parameters that increase the damping ratio, increase the open-loop gain, and decrease the steady-state error based on the transfer function of the circuit model that reflects the real system after obtaining the circuit model. These parameters are then used to control the main magnet power supply circuit, thereby improving the performance of the main magnet power supply in the circuit. The low-precision main magnet power supply information refers to the low-precision main magnet power supply information caused by electromagnetic and radiation interference in the near-field environment. The high-precision main magnet power supply information refers to the high-precision main magnet power supply information formed by using network communication to move away from electromagnetic and radiation interference.

[0061] Furthermore, the 232 communication to network communication unit is sequentially configured with: a 232 to network communication subunit, an equipment-side Ethernet switch subunit, a control room Ethernet switch subunit, a main control room real-time analysis subunit, and a current measurement data subunit. The 232 to network communication subunit includes a DCCT, a high-precision multimeter, and a virtual serial port module. High-precision current measurement values ​​are obtained through the DCCT and high-precision multimeter, and then connected to the virtual serial port module, which in turn connects to the network switch, enabling the high-precision current measurement values ​​to be transmitted via Ethernet. This extends the transmission distance and reduces interference. The main control room real-time analysis subunit transmits the high-precision current data to the main control computer in real time. The main control computer's analysis program then plots the real-time current change graph and analyzes and calculates various performance parameters of the output current.

[0062] Furthermore, the circuit modeling unit based on damping ratio verification is sequentially configured with: a power supply loop modeling subunit, a model damping ratio calculation subunit, a model error control subunit, and a model output subunit. The power supply loop modeling subunit uses simulation software to model the power supply loop between the power source and the superconducting magnet, and outputs the model to the model damping ratio calculation subunit. The model error control subunit compares the calculated model damping ratio with the actual system damping ratio, controlling the error to within 5%, thereby obtaining a circuit model that reflects the real system. The model output subunit sends the model that reflects the real system to the parameter tuning unit based on the damping ratio, open-loop gain, and steady-state error.

[0063] Furthermore, the power supply circuit modeling subunit includes a power controller type acquisition module, a main magnet load parameter acquisition module, and a power supply circuit modeling module. The power supply circuit modeling module uses simulation software to model the power supply circuit containing the power controller type and main magnet load parameters.

[0064] Furthermore, the control model error subunit includes a module for calculating the damping ratio of the actual system, a module for obtaining the damping ratio of the calculated model, a damping ratio error comparison module, and a control model accuracy module. The control model accuracy module determines whether to remodel based on the comparison result of the damping ratio error comparison module. If the error is less than 5%, the established model is considered to reflect the parameters of the real system; otherwise, the power supply circuit modeling subunit will remodel.

[0065] Furthermore, the parameter tuning unit based on damping ratio, open-loop gain, and steady-state error includes a control parameter derivation subunit, a control parameter adjustment subunit, and a control parameter tuning subunit. The control parameter derivation subunit derives the relationship between the control parameters and the model damping ratio, open-loop gain, and steady-state error based on the model's transfer function. The control parameters include PID parameters. The adjustment subunit adjusts the PID parameters according to the derived relationship, appropriately increasing the model damping ratio, appropriately increasing the open-loop gain, and decreasing the steady-state error, thereby completing the parameter tuning. The control parameter tuning subunit performs output current simulation on the tuned parameters and observes the simulation effect. If the performance and system stability meet the set requirements, the tuned parameters are input into the main magnet power supply.

[0066] Furthermore, the appropriate increase in model damping ratio, appropriate increase in open-loop gain, and reduction in steady-state error are specifically as follows: The principle for appropriately increasing the damping ratio is that, after the increase, the response characteristics of the output current in the simulation results should not become overdamped, and the current overshoot should not completely disappear; the principle for appropriately increasing the open-loop gain is that, after the increase, the stability of the model should not be negative when using methods such as the Routh criterion to evaluate it; simultaneously, the oscillation settling time should not increase in the simulation results of the model output current; the principle for reducing the steady-state error is that the steady-state error of the model should be reduced to 0 as much as possible, and if it cannot be reduced to 0, it should be minimized as much as possible, and the overall picture should be considered, ensuring that the damping ratio of the model increases while reducing the steady-state error.

[0067] Furthermore, the performance and system stability meet the set requirements, specifically: based on the difference between the output current stability and the target value in the analyzed power supply performance, if the difference is within 5 times, the output current stability of the model after parameter tuning should be increased by the corresponding factor; if the difference is more than 5 times, the output current stability of the model after parameter tuning should be increased by at least 5 times; at the same time, the output current of the model after parameter tuning should remain stable and should not be out of control.

[0068] A method for optimizing the power supply performance of a superconducting main magnet in an accelerator, such as... Figure 5 As shown, its characteristics include the following steps:

[0069] Step 1: Convert the communication method of the main magnet power supply circuit, which is close-range and susceptible to electromagnetic interference, into a long-range and interference-resistant main magnet network communication method, and obtain high-precision current information of the remote main magnet power supply circuit;

[0070] Step 2: Circuit modeling based on high-precision current information and damping ratio verification: Calculate the model damping ratio and the actual system damping ratio based on the acquired high-precision current data. Compare the two as a criterion for whether the model can reflect the actual circuit system. If the difference is less than 5%, the model is considered to reflect the actual circuit system; otherwise, the model should be remodeled.

[0071] Step 3: Parameter tuning based on damping ratio, open-loop gain and steady-state error: Adjust the control parameters in the obtained circuit model, and judge whether the performance has been improved through simulation analysis, so as to obtain better control parameters.

[0072] Furthermore, adjusting the control parameters in the obtained circuit model specifically involves: appropriately increasing the model damping ratio, appropriately increasing the model open-loop gain, and reducing the model steady-state error, thereby completing the parameter tuning; and performing current simulation on the tuned parameters and observing the simulation effect. If the performance and system stability meet the set requirements, the tuned parameters are then input into the main magnet power supply.

[0073] Example 1: Circuit Modeling Unit Based on Damping Ratio Verification

[0074] Take the power supply circuit of a superconducting main magnet of an accelerator as an example.

[0075] The first step is to obtain the controller type and parameters of the power supply based on the main magnet power supply: the superconducting main magnet power supply controller is PID control, and the initial values ​​of the PID parameters are: P parameter is 20, I parameter is 1, and D parameter is 0; the main magnet load parameters are obtained based on the main magnet load: the inductance of the main magnet when it is in rated operation is 118H, and the total resistance of the power supply circuit is 0.015 ohms.

[0076] The second step is to use simulation software to model the power supply circuit. The modeling objects include the power controller type and parameters mentioned above, and the main magnet load parameters.

[0077] The third step is to calculate the damping ratio of the model. The calculation method is as follows: First, derive the closed-loop transfer function G(s) of the model and transform it into "tail 1" form. The derived closed-loop transfer function G(s) is:

[0078]

[0079] The characteristic equation of the system is:

[0080] 118s 2 +20.015s+1=0(2)

[0081] The damping ratio of the model can be obtained by solving the system of equations based on the damping ratio calculation formula, where ξ is the damping ratio and ω is the damping ratio. n The damping ratio is calculated using the following formula, where the frequency is the damping oscillation frequency:

[0082] When the system characteristic equation is as 2 When +bs+1=0:

[0083] When the system characteristic equation is as 2 When +bs+1=0:

[0084]

[0085] Then, according to formulas (2) and (3), we can obtain the system of equations:

[0086]

[0087] The calculated damping ratio of the model is 0.9213.

[0088] ω n =0.0921

[0089] ξ=0.9213(5)

[0090] The fourth step is to calculate the damping ratio of the actual system based on the acquired current measurement data. The calculation process is as follows: Based on the current monitoring value, the overshoot σ% can be calculated. The calculation method is (peak current - steady-state average current) / (steady-state average current). The measured peak current is 244.2381A, and the steady-state average current is 243.7293A. Therefore, the overshoot is calculated as follows:

[0091]

[0092] Based on the relationship between overshoot and system damping ratio, the damping ratio of the actual system can be calculated. The relationship formula is as follows, and the damping ratio of the actual system can be calculated to be 0.8913.

[0093]

[0094] The fifth step involves comparing the calculated model damping ratio with the actual system damping ratio: the damping ratio of the established model is 0.9213, while the actual system damping ratio is 0.8913, a difference of 0.03. This represents a deviation of 3.37% from the actual damping ratio, which is within 5%. Therefore, the established model can be considered close to the actual system. This yields a circuit model that reflects the actual system.

[0095] Example 2 - Parameter Tuning Unit Based on Damping Ratio, Open-Loop Gain, and Steady-State Error

[0096] Take the power supply circuit of a superconducting main magnet in an accelerator as an example. Figure 6a , Figure 6b As shown in Figure 1, the monitored power supply output current data shows a current stability of 87.33 ppm, far exceeding the required 10 ppm. Moreover, when the current rises, the oscillation adjustment time after overshoot is as long as 851 seconds, exceeding 14 minutes, requiring performance optimization. This power supply circuit has obtained a circuit model that reflects the actual system through the "modeling unit based on damping ratio verification".

[0097] The first step, based on the model, is to derive the relationship between the parameters and the damping ratio, open-loop gain, and steady-state error:

[0098] (1) Derivation of the relationship between parameters and damping ratio: The parameter values ​​of the proportional element P and the integral element I in the model are represented by the letters P and I. The parameter D is not suitable for this system and will cause high-frequency oscillations, so it is kept at the initial value of 0 and is not considered. The transfer function of the model is:

[0099]

[0100] The characteristic equation of the system is:

[0101] 118s 2+(P+0.015)s+I=0(9)

[0102] The damping ratio can be calculated using the formula, which yields a set of equations. These equations can then be used to solve for the damping ratio of the model, where ξ is the damping ratio and ω... n The damping ratio is calculated using the following formula, where the frequency is the damping oscillation frequency:

[0103] When the system characteristic equation is as 2 When +bs+1=0:

[0104]

[0105] Then, according to formulas (2) and (3), we can obtain the system of equations:

[0106]

[0107] Furthermore, the relationship between the damping ratio and the P and I parameters can be derived as follows:

[0108]

[0109] (2) Then, the relationship between the P-parameters, I-parameters, and the open-loop gain of the model is derived: The open-loop transfer function K(s) of the model is:

[0110]

[0111] Equalize the numerator and denominator of the transfer function into the form of multiplication of "tail-one" polynomials:

[0112]

[0113] Then its open-loop gain is This yields the relationship between the system's open-loop gain and the control parameters.

[0114] (3) Finally, the relationship between the control parameters and static error in the model is studied: Since the system uses PID control, according to the basic principle of PID control, the static error of the output is mainly related to the I parameter. The I parameter corresponds to the integral element in PID control, which is controlled according to the difference between the actual value and the target value of the output current, and finally achieves the effect of minimizing the difference between the actual value and the target value. The larger the I parameter, the stronger the effect of the integral element, and the smaller the static error of the model.

[0115] The second step is to adjust the parameters based on the relationship: adjust the parameters according to the relationship between the parameters derived in the first step and the damping ratio, open-loop gain and static error.

[0116] (1) Based on the relationship between the P and I parameters and the damping ratio, it can be seen that the damping ratio of the model is directly proportional to the value of parameter P and inversely proportional to the value of parameter I. Therefore, the damping ratio can be increased by increasing the value of P and decreasing the value of I. However, after increasing the P parameter, it is necessary to check whether the increase is appropriate through simulation results. If the output current becomes overdamped and the overshoot disappears, it proves that the P parameter has been increased too much and is inappropriate.

[0117] (2) Based on the relationship between the P and I parameters and the open-loop gain, it can be seen that the open-loop gain can be increased by increasing the value of I. However, it is necessary to check whether the increase is appropriate through simulation results. In the simulation results of the model output current, the oscillation settling time after overshoot should not increase, and the system should not become unstable.

[0118] (3) Based on the relationship between the I parameter and the steady-state error, it can be seen that the steady-state error can be reduced by increasing the value of the I parameter. However, increasing the I parameter should not reduce the overall damping ratio of the system, otherwise the optimization effect will not be achieved.

[0119] In summary, the adjustment directions for the P and I parameters were determined: the P parameter should be increased; the I parameter should also be increased moderately, but not so much as to decrease the model damping ratio (compared to before parameter tuning). After multiple parameter adjustments and iterations, a better set of parameters was finally obtained, increasing the P parameter from 20 to 60 and the I parameter from 1 to 5.

[0120] The third step is to perform output current simulation after parameter tuning: the simulation results before and after parameter tuning are shown in the figure below. Figure 6c , 6d As shown in Figures 6e and 6f, it can be seen that after parameter tuning, the oscillation settling time after current overshoot is significantly shortened, and the steady-state fluctuation is reduced from a maximum of 0.56A before parameter tuning to a maximum of 0.11A after parameter tuning, a reduction of 5.1 times. Furthermore, the current simulation results do not show any system instability. This proves that the parameter tuning is effective. The tuned parameters are then input into the power supply.

[0121] After applying this set of parameters to the actual system, the current stability was significantly improved. A comparison graph of the actual system is shown below. Figure 6a , 6g As shown in 6h. From Figure 6a and Figure 6h The comparison shows that the output current stability decreased from 87.33ppm to 8.30ppm, which is sufficient to meet the target of 10ppm. Furthermore, from... Figure 6g It can be seen that the overshoot of the power supply output current has been significantly reduced, and the oscillation settling time after overshoot has also been significantly reduced, from 851s to 221s. Performance optimization has been completed.

[0122] The above description is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined in the claims, they should all fall within the protection scope of the present invention.

Claims

1. A power supply performance optimization system for a superconducting main magnet in an accelerator, comprising a 232 communication-to-network communication unit, a circuit modeling unit based on damping ratio verification, and a parameter tuning unit based on damping ratio, open-loop gain, and steady-state error; characterized in that: The 232 communication-to-network communication unit is used to convert the short-range, electromagnetically susceptible communication method into a long-range, interference-resistant network communication method, so as to remotely obtain high-precision main magnet power supply information and send the high-precision current measurement data of the main magnet power supply to the circuit modeling unit based on damping ratio verification. The circuit modeling unit based on damping ratio verification is used to simulate and model the power supply circuit of the superconducting magnet based on the high-precision current information and the main magnet power supply circuit information, to obtain a circuit model that can reflect the real system, and to send the circuit model to the parameter tuning unit based on damping ratio, open-loop gain and steady-state error. The parameter tuning unit based on damping ratio, open-loop gain, and steady-state error is used to derive tuning parameters that increase the damping ratio, increase the open-loop gain, and decrease the steady-state error based on the transfer function of the circuit model that reflects the real system after obtaining the circuit model. These parameters are then used to control the main magnet power supply circuit, thereby improving the performance of the main magnet power supply in the circuit. The high-precision main magnet power supply information is generated by using network communication to avoid electromagnetic and radiation interference. The circuit modeling unit based on damping ratio verification is sequentially configured with: a power supply loop modeling subunit, a model damping ratio calculation subunit, a model error control subunit, and a model output subunit. The power supply loop modeling subunit uses simulation software to model the power supply loop between the power source and the superconducting magnet, and outputs the model to the model damping ratio calculation subunit. The model error control subunit compares the calculated model damping ratio with the actual system damping ratio, controlling the error to within 5%, thereby obtaining a circuit model that reflects the real system. The model output subunit sends the model reflecting the real system to the parameter tuning unit based on the damping ratio, open-loop gain, and steady-state error. The parameter tuning unit based on damping ratio, open-loop gain, and steady-state error includes a control parameter derivation subunit, a control parameter adjustment subunit, and a control parameter tuning subunit. The control parameter derivation subunit derives the relationship between the control parameters and the model's damping ratio, open-loop gain, and steady-state error based on the model's transfer function. The control parameters include PID parameters. The control parameter adjustment subunit adjusts the PID parameters according to the derived relationship, appropriately increasing the model's damping ratio, appropriately increasing the open-loop gain, and decreasing the steady-state error, thereby completing the parameter tuning. The control parameter tuning subunit performs output current simulation on the tuned parameters and observes the simulation effect. If the performance and system stability meet the set requirements, the tuned parameters are input to the main magnet power supply.

2. The accelerator superconducting main magnet power supply performance optimization system according to claim 1, characterized in that: The RS-232 communication to network communication unit is sequentially configured with: a RS-232 to network communication subunit, an equipment-side Ethernet switch subunit, a control room Ethernet switch subunit, a main control room real-time analysis subunit, and a current measurement data subunit. The RS-232 to network communication subunit includes a DCCT, a high-precision multimeter, and a virtual serial port module. High-precision current measurements are obtained through the DCCT and high-precision multimeter, then connected to the virtual serial port module, which in turn connects to the network switch, enabling the high-precision current measurement values ​​to be transmitted via Ethernet. This extends the transmission distance and reduces interference. The main control room real-time analysis subunit transmits the high-precision current data to the main control computer in real time. The main control computer's analysis program then plots real-time current changes and analyzes and calculates various performance parameters of the output current.

3. The accelerator superconducting main magnet power supply performance optimization system according to claim 1, characterized in that: The power supply circuit modeling subunit includes a power controller type acquisition module, a main magnet load parameter acquisition module, and a power supply circuit modeling module. The power supply circuit modeling module uses simulation software to model the power supply circuit containing the power controller type and main magnet load parameters.

4. The accelerator superconducting main magnet power supply performance optimization system according to claim 1, characterized in that: The control model error subunit includes a module for calculating the damping ratio of the actual system, a module for obtaining the damping ratio of the calculated model, a damping ratio error comparison module, and a control model accuracy module. The control model accuracy module determines whether to remodel based on the comparison result of the damping ratio error comparison module. If the error is less than 5%, the established model is considered to reflect the parameters of the real system; otherwise, the power supply circuit modeling subunit will remodel.

5. The accelerator superconducting main magnet power supply performance optimization system according to claim 1, characterized in that: The appropriate increase in model damping ratio, appropriate increase in open-loop gain, and reduction in steady-state error are specifically as follows: The principle for appropriately increasing the damping ratio is that, in the simulation results of the output current of the model after the increase, the response characteristics of the output current should not become overdamped, and the current overshoot should not completely disappear; The principle for appropriately increasing the open-loop gain is: after increasing it, use methods such as the Routh criterion to evaluate the model stability, and the stability should not be negative; at the same time, the oscillation settling time should not increase in the simulation results of the model output current; the principle for reducing the steady-state error is: the steady-state error of the model should be reduced to 0 as much as possible, and if it cannot be reduced to 0, it should be reduced as much as possible, and the overall situation should be considered. While reducing the steady-state error, the damping ratio of the model should be increased.

6. The accelerator superconducting main magnet power supply performance optimization system according to claim 1, characterized in that: The performance and system stability meet the set requirements, specifically: based on the difference between the output current stability and the target value in the analyzed power supply performance, if the difference is within 5 times, the output current stability of the model after parameter tuning should be increased by the corresponding factor; if the difference is more than 5 times, the output current stability of the model after parameter tuning should be increased by at least 5 times; at the same time, the output current of the model after parameter tuning should remain stable and should not be out of control.

7. A method for optimizing the power supply performance of an accelerator superconducting main magnet, based on any one of claims 1-6, characterized in that, Includes the following steps: Step 1: Convert the communication method of the main magnet power supply circuit, which is close-range and susceptible to electromagnetic interference, into a long-range and interference-resistant main magnet network communication method, and obtain high-precision current information of the remote main magnet power supply circuit; Step 2: Circuit modeling based on high-precision current information and damping ratio verification: Calculate the model damping ratio and the actual system damping ratio based on the acquired high-precision current data. Compare the two as a criterion for whether the model can reflect the actual circuit system. If the difference is less than 5%, the model is considered to reflect the actual circuit system; otherwise, the model should be remodeled. Step 3: Parameter tuning based on damping ratio, open-loop gain and steady-state error: Adjust the control parameters in the obtained circuit model, and judge whether the performance has been improved through simulation analysis, so as to obtain better control parameters.

8. The method for optimizing the power supply performance of a superconducting main magnet in an accelerator according to claim 7, characterized in that, The adjustment of control parameters in the obtained circuit model specifically involves: appropriately increasing the model damping ratio, appropriately increasing the model open-loop gain, and reducing the model steady-state error, thereby completing the parameter tuning; and performing current simulation on the tuned parameters and observing the simulation effect. If the performance and system stability meet the set requirements, the tuned parameters are then input into the main magnet power supply.

Images