Multi-objective optimized green insulating gas and solid material configuration method and system

Abstract

The invention relates to the technical field of electrical engineering and high-voltage insulation, discloses a multi-objective optimized green insulating gas and solid material configuration method and system, and aims to solve the problems of nanometer air gap accumulation and insulation failure caused by asynchronous static matching and control of a gas-solid interface under a dynamic electric field. According to the method, nanoscale strain and temperature data are obtained in real time through an interface state high-precision sensing module, electric field gradient information is combined, the data are input into a multi-target collaborative optimization control module based on a Hamiltonian mechanical model, and a 100-nanosecond-level synchronous solid pre-tightening force and gas pressure regulation and control instruction is generated with the purpose of minimizing the system energy change rate; and the piezoelectric actuator and the micro valve array are used for executing microsecond-level dynamic adjustment. According to the system, ultrahigh temporal-spatial resolution monitoring, hardware-level synchronous regulation and control and predictive control of physical model driving are realized, the reliability of green insulation equipment is remarkably improved, and the service life of the green insulation equipment is remarkably prolonged.

Description

Technical Field

[0001] This invention belongs to the field of electrical engineering and high-voltage insulation technology, specifically relating to a multi-objective optimized method and system for configuring green insulating gas and solid materials. Background Technology

[0002] As the global carbon neutrality strategy deepens, power systems are accelerating the phasing out of sulfur hexafluoride (SF6) insulating gases, which have high global warming potential, and shifting towards green alternatives. While these environmentally friendly gases significantly reduce environmental impact, they have revealed serious compatibility defects with solid epoxy resin interfaces under dynamic electric fields in the actual operation of high-voltage gas-insulated switchgear. Existing technologies generally treat the gas-solid interface as a static rigid connection, employing fixed concentration ratios and independent mechanical pre-tightening strategies based on standards such as IEC 61936-1:2010. This neglects the dynamic disturbance mechanism of the interface's nanoscale structure caused by microsecond-level electric field gradient abrupt changes induced by lightning surges or operational overvoltages.

[0003] The reliability of green insulation systems is highly dependent on the stress coordination capability of the gas-solid interface under extreme conditions. Current mainstream solutions attempt to maintain interface integrity by pre-setting gas pressure or constant preload, but their control logic severs the dynamic coupling relationship between gas compressibility and Young's modulus of the solid material. When the electric field changes transiently, the interface generates air gaps of 5–50 nm due to stress mismatch, becoming a major cause of partial discharge and insulation failure. Industry practice shows that simply increasing the gas concentration or mechanical preload strength not only fails to suppress air gap expansion but may also exacerbate material fatigue due to overcompensation, leading to increased life-cycle costs.

[0004] Based on this, the question is how to achieve hardware-level closed-loop adaptive coordination of gas-solid interface stress under dynamic electric field abrupt changes, so as to suppress the cumulative expansion of nano-gaps and ensure the long-term reliability of green insulation systems. Summary of the Invention

[0005] To address the problem of cumulative expansion of nanoscale air gaps and insulation failure at the interface in existing green insulating gas and solid insulating material combination systems under dynamic electric field stress due to static proportioning, misalignment of independent control timing, and insufficient microscopic monitoring accuracy, this invention provides a multi-objective optimized method and system for configuring green insulating gas and solid materials. This invention constructs a closed-loop adaptive control system integrating high-frequency sensing of nanoscale interface deformation, predictive control based on a physical model, and hardware-level synchronous drive. This system suppresses stress mismatch at the gas-solid interface under abrupt changes in electric field gradient in real time, thereby ensuring the long-term reliability of power equipment while meeting multiple constraints such as insulation strength, environmental friendliness, and economy.

[0006] The technical problem to be solved by the present invention is to provide a technical solution that can control the pressure of green insulating gas components and the pre-tightening force of solid insulating components in real time, accurately and synergistically, so as to actively eliminate the generation and development of interfacial nano-gaps under dynamic working conditions.

[0007] To achieve the above objectives, the present invention provides a multi-objective optimized method for configuring green insulating gas and solid materials, which includes the following steps: By setting a high-precision interface state sensing module in the gas-solid insulation interface region, multi-dimensional dynamic physical state data of the interface can be acquired in real time. The multi-dimensional dynamic physical state data includes interface normal strain distribution data, interface tangential strain distribution data, strain change rate data, and local temperature distribution data. Multi-dimensional dynamic physical state data and real-time electric field gradient change rate data obtained through electric field sensors are input into the multi-objective collaborative optimization control module. The Hamiltonian mechanical predictive control model embedded in the multi-objective collaborative optimization control module calculates the Hamiltonian function value and its time rate of change, which characterize the total energy of the gas-solid insulating interface system, in real time based on the input multi-dimensional dynamic physical state data and real-time electric field gradient change rate data. The Hamiltonian mechanical predictive control model optimizes by minimizing the time change rate of the Hamiltonian function value. It generates a set of hardware-level synchronous control commands, including a solid preload control signal sequence and a gas component pressure control signal sequence. The two signal sequences have a phase difference of less than 100 nanoseconds on the time base. The hardware-level synchronous control command is sent to the dynamic collaborative control execution module. The dynamic collaborative control execution module drives the piezoelectric stack actuator array according to the solid preload control signal sequence to apply dynamic preload to the solid insulating component, and drives the piezoelectric micro-valve array according to the gas component pressure control signal sequence to perform microsecond-level fine adjustment of the gas component concentration and total pressure in the insulating gas chamber.

[0008] As one embodiment of the present invention, acquiring multi-dimensional dynamic physical state data of the interface through a high-precision interface state perception module specifically includes: The high-precision interface state sensing module includes a piezoelectric transducer array and a laser Doppler heterodyne interferometer arranged along a preset path of the gas-solid insulating interface. A central clock signal source synchronously triggers the excitation end transducer in the piezoelectric transducer array to generate surface acoustic waves on the surface of a solid insulating component at a center frequency of 100 MHz. Surface acoustic waves propagate along a predetermined path, and their waveforms are modulated in phase and frequency due to nanoscale deformation of the interface in the normal and tangential directions. A laser Doppler heterodyne interferometer emits a probe laser beam onto the surface of a solid insulating component and receives the reflected laser beam modulated by surface acoustic waves. By analyzing the interference signal formed by the reflected laser and the reference laser, the phase and frequency offset of the surface acoustic waves distributed along the path are demodulated, thereby retrieving interface strain distribution data and strain rate data with a spatial resolution of one nanometer and a time sampling rate of ten megahertz.

[0009] As one embodiment of the present invention, the multi-objective cooperative optimization control module is a hardware control core based on a field-programmable gate array (FPGA), and the Hamiltonian mechanics prediction control model is integrated and solidified into the hardware logic circuit inside the FPGA. The Hamiltonian function is defined as the sum of the elastic potential energy, kinetic energy, internal gas energy, and electric field coupling energy of the interface system. Its specific mathematical expression is: , where σ is the stress tensor, ε is the strain tensor, ρ is the density of the solid material, v is the particle velocity, P is the gas pressure, T is the temperature, and E is the electric field strength; The solver logic circuit inside the field-programmable gate array predicts the system state at the next time step by performing real-time numerical integration of the Hamiltonian canonical equation, and calculates the increment of solid preload and gas pressure required to make dH / dt approach zero based on the prediction result, thereby generating a control signal.

[0010] As one embodiment of the present invention, the generation and transmission of hardware-level synchronous control commands are driven by the same phase-locked loop clock unit inside the field-programmable gate array. The phase-locked loop clock unit provides a unified and highly stable clock reference for the piezoelectric transducer excitation signal of the interface state high-precision sensing module, the piezoelectric stacked actuator drive signal of the dynamic collaborative control execution module, and the piezoelectric micro-valve drive signal, thereby ensuring the delay determinism and execution synchronization of the entire link from sensing to execution.

[0011] As one embodiment of the present invention, the piezoelectric stacked actuator array in the dynamic collaborative control execution module is made of lead zirconate titanate ceramic material and is evenly distributed circumferentially along the support structure of the solid insulating component. Each actuator has an adjustable stroke of 0 to 50 micrometers and a driving force of more than 500 Newtons, with a response time of less than 2 milliseconds. The piezoelectric microvalve array is integrated into the gas circulation pipeline. Each microvalve is a cantilever beam structure manufactured based on microelectromechanical systems technology. Its switching time is less than 1 millisecond. By precisely controlling its opening duty cycle, the precise ratio and injection of nitrogen in the main gas storage tank and tetrafluorodinitrile-based mixed gas in the conditioning gas storage tank can be achieved, thereby dynamically adjusting the gas composition and pressure in the gas chamber.

[0012] To achieve the above objectives, the present invention also provides a multi-objective optimized green insulating gas and solid material preparation system, comprising: The high-precision interface state sensing module is set in the gas-solid insulation interface area to acquire multi-dimensional dynamic physical state data of the interface in real time. The data includes interface normal strain distribution, tangential strain distribution, strain change rate and local temperature distribution. The multi-objective collaborative optimization control module is electrically connected to the interface state high-precision sensing module. It is used to receive multi-dimensional dynamic physical state data and external input real-time electric field gradient change rate data. Based on the internally solidified Hamiltonian mechanical prediction control model, it generates hardware-level synchronized solid preload control signal sequence and gas component pressure control signal sequence with the goal of minimizing the time change rate of the total energy of the interface system. The dynamic collaborative control execution module, which is electrically connected to the multi-objective collaborative optimization control module, is used to receive and execute hardware-level synchronous control commands. This module includes a piezoelectric stacked actuator array for adjusting the preload of solid insulators and a piezoelectric micro-valve array for adjusting the composition and pressure of insulating gas.

[0013] As one embodiment of the present invention, the high-precision interface state sensing module is composed of a surface acoustic wave generation and detection subsystem. The surface acoustic wave generation and detection subsystem includes a piezoelectric transducer array arranged along a preset path of the gas-solid interface, a laser Doppler heterodyne interferometer for emitting and receiving lasers, and a high-speed data processing unit for signal demodulation. The module is able to acquire interface dynamic strain data with a spatial resolution of one nanometer at a sampling rate of ten megahertz.

[0014] As one embodiment of the present invention, the core processing device of the multi-objective collaborative optimization control module is a field-programmable gate array (FPGA). The Hamiltonian mechanics prediction control model is implemented in a hardware description language and synthesized into a parallel processing logic unit and a dedicated numerical calculation unit within the FPGA. The FPGA integrates a phase-locked loop (PLL) clock unit, which provides a high-frequency clock signal that is from the same source and synchronized with the interface state high-precision sensing module and the dynamic collaborative control execution module.

[0015] As one embodiment of the present invention, the piezoelectric stacked actuator array in the dynamic collaborative control execution module is composed of multiple independent piezoelectric stacked actuator units. The actuator units are evenly distributed along the circumference of the solid insulating flange and directly act on the preload bolts or pressure rings. The driving voltage signal is generated by the solid preload control signal output by the multi-objective collaborative optimization control module after high voltage amplification.

[0016] As one embodiment of the present invention, the system further includes a gas component online adjustment unit, which is fluidly connected to the piezoelectric microvalve array in the dynamic collaborative control execution module. The gas component online adjustment unit includes a main gas storage tank for storing nitrogen, a conditioning gas storage tank for storing tetrafluorodinitrile-based mixed gas, a set of high-precision mass flow controllers, and a gas mixing chamber. The piezoelectric microvalve array precisely controls the gas flow rate entering the mixing chamber from the two storage tanks according to the gas component pressure control signal sequence, so as to realize the dynamic active adjustment of the dielectric constant and breakdown field strength of the gas in the insulating chamber.

[0017] In summary, this application includes at least one of the following beneficial technical effects: First, it achieves ultra-high precision real-time monitoring of the gas-solid insulation interface state. By employing a sensing module based on surface acoustic wave and laser Doppler heterodyne interferometry, the spatial resolution of interface deformation monitoring is improved to the nanometer level, and the temporal resolution is improved to the sub-microsecond level. This completely solves the shortcomings of existing strain gauges or fiber optic grating sensors in terms of spatial resolution and sampling frequency, providing unprecedented high-fidelity real-time data input for closed-loop control.

[0018] Second, a hardware-level synchronous and coordinated control mechanism was constructed. By utilizing the parallel processing capability of the field-programmable gate array and the high synchronization of the built-in phase-locked loop clock, the uncertainty and delay in control timing caused by the traditional microprocessor-based and software polling mechanism were fundamentally eliminated. This ensured that the phase difference between the two actions of solid preload compensation and gas pressure compensation was strictly controlled within one hundred nanoseconds, achieving instantaneous and synchronous response to dynamic electric field stress.

[0019] Third, a predictive control strategy based on a physical model was introduced. The Hamiltonian mechanical predictive control model was used to replace the traditional proportional-integral-derivative controller, elevating control decision-making from simple error feedback to a profound understanding and proactive intervention of the energy state evolution of the interface system. This allows the control system to predict and suppress the initiation of nanoscale air gaps in advance, rather than passively compensating after their formation. This achieves feedforward, optimal control of the dynamic stability of the insulation system, significantly improving the operational reliability and service life of green insulated power equipment. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the multi-objective optimized green insulating gas and solid material configuration method of the present invention; Figure 2 This is a schematic diagram of the configuration system in this invention. Detailed Implementation

[0021] This invention provides a multi-objective optimized method and system for configuring green insulating gas and solid materials. It aims to solve the problem of cumulative expansion of nanoscale air gaps and insulation failure at the interface in existing green insulating gas and solid insulating material combination systems under dynamic electric field stress, caused by static proportioning, misalignment of independent control timing, and insufficient microscopic monitoring accuracy. This method constructs a closed-loop adaptive control system integrating high-frequency sensing of nanoscale interface deformation, predictive control based on a physical model, and hardware-level synchronous drive. This system suppresses stress mismatch at the gas-solid interface under abrupt changes in electric field gradient in real time, thereby ensuring the long-term reliability of power equipment while meeting multiple constraints such as insulation strength, environmental friendliness, and economy.

[0022] Combined with appendix Figures 1 to 2 In this application, the multi-objective optimized method for configuring green insulating gas and solid materials includes the following steps: S1, through the high-precision interface state sensing module set in the gas-solid insulation interface area, real-time acquisition of multi-dimensional dynamic physical state data of the interface, including interface normal strain distribution data, interface tangential strain distribution data, strain change rate data and local temperature distribution data. S2 inputs the multi-dimensional dynamic physical state data and the real-time electric field gradient change rate data obtained through the electric field sensor into the multi-objective collaborative optimization control module. S3, the Hamiltonian mechanical predictive control model embedded in the multi-objective collaborative optimization control module, calculates the Hamiltonian function value and its time rate of change in the gas-solid insulating interface system in real time based on the input multi-dimensional dynamic physical state data and real-time electric field gradient change rate data. S4, the Hamiltonian mechanical prediction and control model optimizes by minimizing the time change rate of the Hamiltonian function value and generates a set of hardware-level synchronous control instructions. The synchronous control instructions include a solid preload control signal sequence and a gas component pressure control signal sequence. The two signal sequences have a phase difference of less than 100 nanoseconds on the time base. S5 sends hardware-level synchronous control commands to the dynamic collaborative control execution module. The dynamic collaborative control execution module drives the piezoelectric stack actuator array according to the solid preload control signal sequence to apply dynamic preload to the solid insulating component, and drives the piezoelectric micro-valve array according to the gas component pressure control signal sequence to perform microsecond-level fine adjustment of the gas component concentration and total pressure in the insulating gas chamber.

[0023] In step S1, the high-precision interface state sensing module includes a piezoelectric transducer array and a laser Doppler heterodyne interferometer arranged along a preset path along the gas-solid insulating interface. This module is synchronously triggered by a central clock signal source to ensure that all sensing units operate under a unified time reference.

[0024] Specifically, the excitation transducer in the piezoelectric transducer array is triggered by a central clock signal source at a center frequency of 100 MHz, generating surface acoustic waves on the surface of a solid insulator. These surface acoustic waves propagate along a predetermined path, and their propagation characteristics are modulated by the nanoscale deformation of the interface, manifesting as phase shift and frequency drift.

[0025] A laser Doppler heterodyne interferometer emits a probe laser beam onto the surface of a solid insulating component and receives the reflected laser beam modulated by surface acoustic waves. The reflected laser beam and the reference laser beam form a heterodyne interference signal inside the interferometer. This signal is converted into an electrical signal by a high-speed photodetector and then sent to a high-speed data processing unit for demodulation.

[0026] The high-speed data processing unit extracts the surface acoustic wave phase and frequency offset distributed along the path from the interference signal using Fourier transform and time-frequency analysis algorithms. It further inversely derives the interface normal strain distribution, tangential strain distribution, strain rate of change, and local temperature distribution data acquired through an auxiliary temperature sensor. The spatial resolution of the entire sensing process reaches one nanometer, and the temporal sampling rate reaches ten megahertz, sufficient to capture the microsecond-level dynamic response of the interface caused by abrupt changes in the electric field gradient (dE / dt > 5 kV / μs).

[0027] In step S2, multi-dimensional dynamic physical state data and real-time electric field gradient change rate data are synchronously transmitted to the multi-objective collaborative optimization control module. The electric field gradient change rate data is acquired by an electric field sensor array arranged inside the insulating cavity. This sensor array uses a high-bandwidth capacitive probe with a response bandwidth of not less than 50 MHz, which can accurately capture the instantaneous changes in electric field intensity in the spatial and temporal dimensions.

[0028] Before entering the multi-objective collaborative optimization control module, all data undergoes timestamp alignment and noise filtering to eliminate errors introduced by transmission delay and electromagnetic interference. Timestamp alignment is performed based on a global time reference provided by a central clock signal source, ensuring strict synchronization between physical state data and electric field data on the time axis.

[0029] In step S3, the core of the multi-objective collaborative optimization control module is a field-programmable gate array (FPGA) containing a Hamiltonian mechanical predictive control model. This model treats the gas-solid insulating interface as an open dissipative system, whose total energy consists of elastic potential energy, kinetic energy, internal gas energy, and electric field coupling energy. The mathematical expression for the Hamiltonian function H is: Where σ is the stress tensor in the solid insulating material, ε is the corresponding strain tensor, ρ is the density of the solid material, v is the particle velocity vector, P is the gas pressure, T is the local temperature, E is the electric field intensity vector, ε0 is the vacuum permittivity, ε_r is the relative permittivity, V_solid is the volume of the solid insulating component, V_gas is the volume of the gas chamber, and V_total is the total volume of the entire insulation system. A dedicated numerical computation unit within the field-programmable gate array (FPGA) discretizes the above integrals and, combined with the input real-time physical state data and electric field gradient rate of change data, calculates the Hamiltonian function value H(t) and its time derivative dH / dt at the current moment in real time. This calculation process employs a parallel pipeline architecture, with a single iteration taking no more than two hundred nanoseconds.

[0030] In step S4, the Hamiltonian mechanical predictive control model aims to minimize dH / dt and solves for the solid preload increment ΔF and gas pressure increment ΔP required for the next control cycle. This solution process is based on the Hamiltonian canonical equations: Where q represents the generalized coordinates (corresponding to the interface displacement field) and p represents the generalized momentum (corresponding to the interface stress field). The solver logic circuit inside the field-programmable gate array (FPGA) performs explicit Euler integration or Runge-Kutta fourth-order integration on the above equations to predict the system state evolution trajectory within the next ten microseconds. Based on this, the optimal control input that makes dH / dt approach zero is calculated using the gradient descent method or the Newton-Raphson iteration method.

[0031] The final generated solid preload control signal sequence is a series of voltage pulses with an amplitude range of 0 to 200 volts and a pulse width resolution of no less than 10 nanoseconds. The gas component pressure control signal sequence is a pulse width modulation signal with an adjustable duty cycle, a frequency of 100 kHz, and a duty cycle adjustment accuracy of 0.1%. Both signal sequences are generated by the same phase-locked loop clock unit, ensuring that the phase difference between their rising and falling edges is strictly controlled within 100 nanoseconds.

[0032] In step S5, the hardware-level synchronous control command is sent to the dynamic collaborative control execution module. This module consists of two parts: a piezoelectric stack actuator array and a piezoelectric microvalve array. The piezoelectric stack actuator array is made of lead zirconate titanate ceramic material, with eight independent units evenly distributed along the circumference of the solid insulating flange. Each unit directly acts on the loading end face of the preload bolt. After receiving the solid preload control signal, the actuator unit amplifies it to a driving voltage of 0 to 500 volts by a high-voltage amplifier, generating an axial displacement of 0 to 50 micrometers, with an output force greater than 500 Newtons and a response time of less than 2 milliseconds.

[0033] The piezoelectric microvalve array is integrated into the gas circulation pipeline and includes two independently controlled microvalves: the first microvalve is connected to the main gas storage tank (filled with high-purity nitrogen), and the second microvalve is connected to the conditioning gas storage tank (filled with tetrafluorodinitrile-based mixed gas, the main components of which are CF3I and N2, GWP<100).

[0034] Each microvalve is a cantilever beam structure manufactured based on microelectromechanical systems (MEMS) technology, with a switching time of less than 1 millisecond. The dynamic collaborative control execution module precisely controls the opening duty cycle of the two microvalves according to the gas component pressure control signal sequence, allowing nitrogen and conditioning gas to flow into the gas mixing chamber in a preset ratio, and then into the insulating chamber. This dynamically adjusts the total gas pressure (ranging from 0.3 to 0.6 MPa) and component concentration (the volume fraction of conditioning gas can be continuously adjusted between 2% and 8%), achieving active control over the gas dielectric strength and breakdown field strength.

[0035] On the other hand, the multi-objective optimized green insulating gas and solid material configuration system disclosed in this application specifically includes a high-precision interface state sensing module, a multi-objective collaborative optimization control module, and a dynamic collaborative regulation execution module. The high-precision interface state sensing module is located in the gas-solid insulation interface region and is used to acquire multi-dimensional dynamic physical state data of the interface in real time.

[0036] The multi-objective collaborative optimization control module is electrically connected to the interface state high-precision sensing module. It is used to receive multi-dimensional dynamic physical state data and external input real-time electric field gradient change rate data. Based on the internally solidified Hamiltonian mechanical predictive control model, it generates hardware-level synchronized solid preload control signal sequence and gas component pressure control signal sequence with the goal of minimizing the time change rate of the total energy of the interface system.

[0037] The dynamic collaborative control execution module is electrically connected to the multi-objective collaborative optimization control module and is used to receive and execute hardware-level synchronous control commands. This module includes a piezoelectric stacked actuator array for adjusting the preload of solid insulators and a piezoelectric micro-valve array for adjusting the composition and pressure of insulating gas.

[0038] The high-precision interface state sensing module consists of a surface acoustic wave generation and detection subsystem, which includes a piezoelectric transducer array arranged along a preset path at the gas-solid interface, a laser Doppler heterodyne interferometer for emitting and receiving lasers, and a high-speed data processing unit for signal demodulation.

[0039] The piezoelectric transducer array employs an interdigital transducer structure with a center frequency of 100 MHz and a bandwidth of 20 MHz. Sixteen sensing nodes are arranged along the circumference of the interface, with a spacing of 5 mm between adjacent nodes. The laser Doppler heterodyne interferometer uses a dual-frequency laser source with a frequency difference of 40 MHz and a detection spot diameter of 50 micrometers, ensuring high sensitivity detection of nanoscale surface displacements.

[0040] The high-speed data processing unit is based on a field-programmable gate array and incorporates a fast Fourier transform kernel and a time-frequency joint analysis algorithm, enabling it to complete a full-path strain field reconstruction every ten microseconds.

[0041] The core processing device of the multi-objective collaborative optimization control module is a Xilinx Ultrascale+ series field-programmable gate array, whose internal logic resources are divided into three functional areas: data preprocessing area, Hamiltonian model calculation area, and control command generation area.

[0042] The data preprocessing area is responsible for denoising, interpolating, and transforming the received raw sensor data; the Hamiltonian model calculation area realizes the discrete integral of the Hamiltonian function and the real-time estimation of dH / dt; the control command generation area performs optimization solutions and outputs synchronous control signals.

[0043] The field-programmable gate array integrates a phase-locked loop clock unit, which outputs a master clock signal with a frequency of 1 MHz. This signal is then distributed through a clock tree network to the excitation circuit of the interface state high-precision sensing module and the drive circuit of the dynamic collaborative control execution module, ensuring that the clock jitter of the entire system is less than ten picoseconds.

[0044] The piezoelectric stack actuator array in the dynamic coordinated control execution module consists of eight independent piezoelectric stack actuator units, each with a height of 30 mm, a diameter of 15 mm, a maximum stroke of 50 micrometers, and a static stiffness of 100 kilonewtons per millimeter.

[0045] The actuator unit is bonded to the metal insert on the back of the solid insulating flange using epoxy resin adhesive, with its output end face in direct contact with the tail of the preload bolt. The high-voltage amplifier employs a broadband piezoelectric drive architecture with a bandwidth of no less than 500 kHz and an output voltage ripple of less than 0.5%. Each microvalve in the piezoelectric microvalve array is fabricated using silicon-based micromachining technology, with a cantilever beam length of 500 micrometers, a width of 50 micrometers, a thickness of 10 micrometers, a drive voltage of 0 to 120 volts, a flow control range of 0 to 10 standard cubic centimeters per minute, and a repeatability better than 0.5%.

[0046] The system also includes an online gas composition blending unit, which is fluidly connected to the piezoelectric microvalve array in the dynamic collaborative control execution module. The online gas composition blending unit includes a main gas tank storing nitrogen, a conditioning gas tank storing a tetrafluorodinitrile-based mixed gas, a set of high-precision mass flow controllers, and a gas mixing chamber. Both the main gas tank and the conditioning gas tank are equipped with pressure and temperature sensors for real-time monitoring of the gas state. The high-precision mass flow controllers have a range of 0 to 50 standard cubic centimeters per minute and a control accuracy of 0.5%. The gas mixing chamber has a turbulence-promoting structure to ensure uniform mixing of the two gases within ten milliseconds. The mixed gas is injected into an insulated chamber via the piezoelectric microvalve array. A gas composition analyzer is installed in the chamber for closed-loop verification of the actual gas composition.

[0047] This embodiment achieves ultra-high precision real-time monitoring of the gas-solid insulation interface state, hardware-level synchronous and coordinated regulation, and predictive control based on a physical model through the above-described method and system.

[0048] It will be apparent to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention. Therefore, the embodiments should be regarded as exemplary and non-limiting in all respects.

[0049] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A multi-objective optimized method for configuring green insulating gas and solid materials, characterized in that, include: By using a high-precision interface state sensing module set in the gas-solid insulation interface region, multi-dimensional dynamic physical state data of the interface can be acquired in real time. Multi-dimensional dynamic physical state data and real-time electric field gradient change rate data obtained through electric field sensors are input into the multi-objective collaborative optimization control module. The Hamiltonian mechanical predictive control model embedded in the multi-objective collaborative optimization control module calculates the Hamiltonian function value and its time rate of change, which characterize the total energy of the gas-solid insulating interface system, in real time based on the input multi-dimensional dynamic physical state data and real-time electric field gradient change rate data. The Hamiltonian mechanical predictive control model optimizes by minimizing the time change rate of the Hamiltonian function value. It generates a set of hardware-level synchronous control commands, including a solid preload control signal sequence and a gas component pressure control signal sequence. The two signal sequences have a phase difference of less than 100 nanoseconds on the time base. The hardware-level synchronous control command is sent to the dynamic collaborative control execution module. The dynamic collaborative control execution module drives the piezoelectric stack actuator array according to the solid preload control signal sequence to apply dynamic preload to the solid insulating component, and drives the piezoelectric micro-valve array according to the gas component pressure control signal sequence to perform microsecond-level fine adjustment of the gas component concentration and total pressure in the insulating gas chamber.

2. The multi-objective optimized method for configuring green insulating gas and solid materials according to claim 1, characterized in that, A high-precision interface state sensing module, positioned at the gas-solid insulation interface region, acquires multi-dimensional dynamic physical state data of the interface in real time, including: A central clock signal source synchronously triggers the excitation end transducer in the piezoelectric transducer array arranged along a preset path at the gas-solid insulation interface to generate surface acoustic waves on the surface of the solid insulation component at a center frequency of 100 MHz. A laser Doppler heterodyne interferometer is used to emit a probe laser onto the surface of a solid insulating component and receive the reflected laser modulated by surface acoustic waves. By analyzing the interference signal formed by the reflected laser and the reference laser, the phase and frequency offset of the surface acoustic waves distributed along the path are demodulated. Based on phase and frequency offsets, interface normal strain distribution data, tangential strain distribution data, strain rate data, and local temperature distribution data obtained by auxiliary temperature sensors are retrieved with a spatial resolution of one nanometer and a time sampling rate of ten megahertz.

3. The multi-objective optimized method for configuring green insulating gas and solid materials according to claim 1, characterized in that, The multi-objective collaborative optimization control module is a hardware control core based on a field-programmable gate array (FPGA). The Hamiltonian mechanical predictive control model is integrated and solidified into the hardware logic circuit inside the FPGA. The Hamiltonian function is defined as the sum of the elastic potential energy, kinetic energy, internal gas energy, and electric field coupling energy of the interface system, and its mathematical expression is: σ is the stress tensor in the solid insulating material, ε is the corresponding strain tensor, ρ is the density of the solid material, v is the particle velocity vector, P is the gas pressure, T is the local temperature, E is the electric field intensity vector, ε_o is the vacuum permittivity, ε_r is the relative permittivity, V_solid is the volume of the solid insulating component, V_gas is the volume of the gas chamber, and V_total is the total volume of the entire insulation system. The solver logic circuit inside the field-programmable gate array predicts the system state at the next time step by performing real-time numerical integration of the Hamiltonian canonical equation. Based on this prediction, it calculates the increment of solid preload and gas pressure required to make dH / dt approach zero, thereby generating a solid preload control signal sequence and a gas component pressure control signal sequence.

4. The multi-objective optimized method for configuring green insulating gas and solid materials according to claim 1, characterized in that, The generation and transmission of hardware-level synchronous control commands are driven by the same phase-locked loop clock unit inside the field-programmable gate array. The phase-locked loop clock unit provides a unified and highly stable clock reference for the piezoelectric transducer excitation signal of the interface state high-precision sensing module, the piezoelectric stack actuator drive signal of the dynamic collaborative control execution module, and the piezoelectric micro-valve drive signal, ensuring the delay determinism and execution synchronization of the entire link from sensing to execution.

5. The multi-objective optimized method for configuring green insulating gas and solid materials according to claim 1, characterized in that, The piezoelectric stacked actuator array in the dynamic coordinated control execution module is made of lead zirconate titanate ceramic material and is evenly distributed circumferentially along the support structure of the solid insulating component. Each actuator has an adjustable stroke of 0 to 50 micrometers and a driving force of more than 500 Newtons, with a response time of less than 2 milliseconds. The piezoelectric microvalve array is integrated into the gas circulation pipeline. Each microvalve is a cantilever beam structure manufactured based on microelectromechanical systems technology. Its switching time is less than 1 millisecond. By precisely controlling its opening duty cycle, the precise ratio and injection of nitrogen in the main gas storage tank and tetrafluorodinitrile-based mixed gas in the conditioning gas storage tank can be achieved, thereby dynamically adjusting the gas composition and pressure in the gas chamber.

6. The multi-objective optimized method for configuring green insulating gas and solid materials according to claim 1, characterized in that, Each actuator unit in the piezoelectric stacked actuator array acts directly on the preload bolts or pressure rings on the solid insulating flange. Its driving voltage signal is generated by the solid preload force regulation signal output by the multi-objective collaborative optimization control module after high voltage amplification.

7. The multi-objective optimized method for configuring green insulating gas and solid materials according to claim 1, characterized in that, Before being input into the multi-objective collaborative optimization control module, the multi-dimensional dynamic physical state data and real-time electric field gradient change rate data undergo timestamp alignment and noise filtering based on the central clock signal source to ensure that the physical state data and electric field data are strictly synchronized on the time axis, and that the response bandwidth of the electric field sensor array is not less than 50 MHz.

8. A multi-objective optimized green insulating gas and solid material configuration system, characterized in that, include: The high-precision interface state sensing module is set in the gas-solid insulation interface area to acquire multi-dimensional dynamic physical state data of the interface in real time. The data includes interface normal strain distribution, tangential strain distribution, strain change rate and local temperature distribution. The multi-objective collaborative optimization control module is electrically connected to the interface state high-precision sensing module. It is used to receive multi-dimensional dynamic physical state data and external input real-time electric field gradient change rate data. Based on the internally solidified Hamiltonian mechanical prediction control model, it generates hardware-level synchronized solid preload control signal sequence and gas component pressure control signal sequence with the goal of minimizing the time change rate of the total energy of the interface system. The dynamic collaborative control execution module, which is electrically connected to the multi-objective collaborative optimization control module, is used to receive and execute hardware-level synchronous control commands. This module includes a piezoelectric stacked actuator array for adjusting the preload of solid insulators and a piezoelectric micro-valve array for adjusting the composition and pressure of insulating gas.

9. The multi-objective optimized green insulating gas and solid material configuration system according to claim 8, characterized in that, The high-precision interface state sensing module consists of a surface acoustic wave generation and detection subsystem, which includes a piezoelectric transducer array arranged along a preset path at the gas-solid interface, a laser Doppler heterodyne interferometer for emitting and receiving lasers, and a high-speed data processing unit for signal demodulation. The core processing device of the multi-objective collaborative optimization control module is a field-programmable gate array (FPGA). The Hamiltonian mechanical predictive control model is implemented in a hardware description language and synthesized into parallel processing logic units and dedicated numerical calculation units within the FPGA. The field-programmable gate array integrates a phase-locked loop clock unit, which provides a high-frequency clock signal that is from the same source and synchronized with the interface state high-precision sensing module and the dynamic collaborative control execution module.

Images