Super-long tunnel sectional type power supply and distribution topological optimization and energy consumption dynamic balancing method

Abstract

The invention discloses an ultra-long tunnel sectional type power supply and distribution topological optimization and energy consumption dynamic balancing method, and relates to the technical field of power systems and automation thereof, and the method comprises the following steps: S1, constructing a tunnel longitudinal sectional type power supply and distribution topological structure, S2, collecting multi-source data in real time, S3, carrying out dynamic energy consumption balancing optimization, and S4, carrying out fault self-healing control. The tunnel is divided into a plurality of independent power supply sections, real-time energy consumption data of lighting equipment, ventilation equipment and monitoring equipment in each power supply section and data such as tunnel traffic flow and environment temperature and humidity are collected in real time, and a dynamic capacity design formula is combined, so that the load balance degree of the transformer is improved by 20%; and a self-adaptive rotation angle mechanism and a catastrophe operator are introduced, so that the convergence rate is optimized by 30%, the network loss is further reduced to 3.1%, and the reactive compensation precision is improved.

Description

Technical Field

[0001] This application belongs to the field of power system and its automation technology, and specifically relates to a method for segmented power supply and distribution topology optimization and dynamic energy consumption balance in ultra-long tunnels. Background Technology

[0002] As a crucial component of modern transportation infrastructure, the stability, safety, and energy efficiency of the power supply and distribution system of ultra-long tunnels directly impact the reliability and economy of tunnel operation. However, existing tunnel power supply and distribution systems generally face the following technical challenges:

[0003] (1) Traditional tunnel power supply and distribution systems adopt a centralized power supply mode, which does not take into account the impact of tunnel length, traffic flow fluctuations and environmental parameter changes on the load, resulting in unbalanced transformer load and large network losses. In addition, the existing system lacks a dynamic optimization mechanism, making it difficult to respond in real time to changes in traffic flow and environmental temperature and humidity fluctuations, resulting in energy waste.

[0004] (2) There are a large number of high-power equipment such as lighting, ventilation, and monitoring inside ultra-long tunnels. Traditional copper cables are susceptible to electromagnetic interference, which can lead to a decrease in signal transmission quality and even cause communication errors or equipment malfunctions. Existing cable materials are mostly pure copper or ordinary alloys, which have insufficient shielding performance and cannot meet the stable power supply requirements in complex electromagnetic environments.

[0005] (3) When a power supply anomaly occurs in a certain section, the traditional system relies on mechanical switches to transfer the load, which has a long response time (≥300ms) and lacks an intelligent fault diagnosis mechanism, making it difficult to quickly locate and isolate the fault point, which can easily cause a large-scale power outage risk. In addition, most existing reactive power compensation devices are fixed and cannot dynamically adjust the compensation parameters, which affects the power supply quality. Summary of the Invention

[0006] This application provides a method for segmented power supply and distribution topology optimization and dynamic energy consumption balancing in ultra-long tunnels to solve the problems existing in the prior art.

[0007] To achieve the above-mentioned objectives, this application provides the following technical solution:

[0008] According to one aspect of this application, a method for segmented power supply and distribution topology optimization and dynamic energy consumption balancing in ultra-long tunnels is provided, comprising the following steps:

[0009] S1. Construct a longitudinal segmented power supply and distribution topology for the tunnel: Divide the tunnel into N power supply sections based on the tunnel length, deploy an independent intelligent power distribution unit in each section, and connect adjacent power distribution units through anti-electromagnetic interference composite cables;

[0010] S2. Real-time acquisition of multi-source data: Real-time energy consumption data of lighting equipment, ventilation equipment and monitoring equipment in each power supply section are acquired through a distributed optical fiber sensor network, while tunnel traffic flow and ambient temperature and humidity data are also collected.

[0011] S3. Dynamic energy consumption balance optimization: Based on the improved quantum genetic algorithm, energy consumption data is modeled, and the parameters of transformer taps and reactive power compensation devices in each power supply section are dynamically adjusted with the goal of minimizing total network loss and load imbalance.

[0012] S4. Fault self-healing control: When the power supply of a certain section is abnormal, a fast switching switch based on graphene composite conductor is activated to automatically transfer the load of the faulty section to the power supply circuit of the adjacent section.

[0013] In one possible implementation, the electromagnetic interference-resistant composite cable adopts a multi-layer shielding structure, and its core wire is made of graphene-modified copper-based composite material. The preparation process includes: ball milling and mixing copper powder and graphene oxide at a mass ratio of 98:2, forming a preform by plasma sintering under argon protection, and then processing it into wire by equal channel angle extrusion process.

[0014] In one possible implementation, the power distribution capacity design of the k-th power supply segment in the segmented power supply topology satisfies the formula:

[0015]

[0016] in For design capacity, For segment length, To predict peak traffic volume, As the baseline capacity factor, This is a dynamic correction factor.

[0017] In one possible implementation, the distributed optical fiber sensor network employs Brillouin optical time-domain reflectometry, with sensor nodes embedded in the tunnel sidewall at a spacing of no more than 50 meters, and its encapsulation shell made of carbon nanotube-reinforced epoxy resin composite material.

[0018] In one possible implementation, the improved quantum genetic algorithm includes:

[0019] The transformer tap position and reactive power compensation are represented by quantum bit encoding;

[0020] Introducing an adaptive rotation angle mechanism, the decay formula for the rotation angle step size Δθ at iteration number t is:

[0021] Δθ_t = θ_0 × e^{-γt}

[0022] Where θ_0 is the initial rotation angle, ranging from 0.1π to 0.3π; γ is the decay coefficient, ranging from 0.02 to 0.05; and t is the current iteration number.

[0023] Add a catastrophe operator to reset 30% of the individual quantum states when the population diversity falls below a threshold.

[0024] In one possible implementation, the preparation of the graphene composite conductor includes the following steps:

[0025] S41. A copper foil is immersed in a 0.5 mol / L silver nitrate solution to carry out a displacement reaction, thereby generating a layer of silver nanoparticles on the surface;

[0026] S42. Vertical graphene arrays were grown on silver nanoparticle layers by chemical vapor deposition to crack methane at 550 °C.

[0027] S43. Perform gradient annealing under a protective atmosphere: heat to 300℃ at 10℃ / min and hold for 1 hour, then heat to 450℃ at 5℃ / min and hold for 2 hours.

[0028] One possible implementation also includes:

[0029] In step S3, a load prediction module is embedded, which uses an LSTM neural network to predict load changes in the next 15 minutes based on historical traffic flow data. The input layer of the prediction model includes feature vectors for date type, time period, and weather status.

[0030] In one possible implementation, each intelligent power distribution unit has a built-in self-diagnostic system that diagnoses winding loosening faults by monitoring the energy change rate in the 2-5kHz frequency band of the transformer vibration spectrum. When the change rate exceeds 15%, an early warning signal is triggered.

[0031] In one possible implementation, the reactive power compensation device adopts a modular multilevel converter structure, and its power module capacitor uses a barium titanate@titanium dioxide core-shell dielectric material, the preparation method of which includes:

[0032] Tetrabutyl titanate is hydrolyzed to generate TiO2 sol;

[0033] A 5-10 nm thick TiO2 layer was coated onto the surface of BaTiO3 nanoparticles via electrostatic self-assembly.

[0034] Dense ceramics are formed by sintering in a nitrogen atmosphere at 1350℃.

[0035] In one possible implementation, a wind-solar hybrid power generation device is deployed at the tunnel entrance, and its output power is connected to the power supply and distribution system through a SiC-MOSFET-based DC-DC converter. The heat dissipation substrate of the converter is made of diamond-aluminum composite metal material with a thermal conductivity ≥600W / (m·K).

[0036] Compared with existing technologies, the segmented power supply and distribution topology optimization and dynamic energy consumption balancing method for ultra-long tunnels described in this application has the following beneficial effects:

[0037] 1. In the above-mentioned method for segmented power supply and distribution topology optimization and dynamic energy consumption balancing in ultra-long tunnels, the tunnel is divided into multiple independent power supply sections, and combined with dynamic capacity design formulas, the transformer load balance is improved by 20%, and network loss is reduced by 12%. By introducing an adaptive rotation angle mechanism and a catastrophe operator, the convergence speed is improved by 30%, network loss is further reduced to 3.1%, and reactive power compensation accuracy is improved.

[0038] 2. This invention improves the conductivity of the cable by using graphene-modified copper-based composite material, which is more than 15% higher than that of traditional copper cables; the multi-layer shielding structure significantly suppresses electromagnetic interference and reduces the communication error rate by 80%.

[0039] 3. This invention utilizes a graphene composite conductor for rapid switching: the response time is ≤50 ms, which is 5 times faster than traditional mechanical switches, improving power supply reliability to 99.99%. The use of barium titanate@titanium dioxide core-shell dielectric material extends capacitor life by 2 times and reduces power loss by 15%.

[0040] 4. This invention uses LSTM neural network for load prediction: combining multi-dimensional feature vectors such as date type, time period, and weather status, the prediction error rate is ≤8%, making transformer capacity design more in line with actual needs and reducing equipment redundancy costs by 10%. By monitoring the energy change rate in the 2-5 kHz frequency band of the vibration spectrum, an early warning can be triggered 30 minutes before the occurrence of winding loosening faults, reducing the risk of sudden shutdowns. Attached Figure Description

[0041] Figure 1 This is a flowchart illustrating the segmented power supply and distribution topology optimization and dynamic energy consumption balance method for ultra-long tunnels proposed in this application. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Any aspects not detailed in this application are well-known technologies to those skilled in the art.

[0043] Example 1:

[0044] Please refer to Figure 1 This embodiment provides a method for segmented power supply and distribution topology optimization and dynamic energy consumption balancing in ultra-long tunnels. The implementation steps are as follows:

[0045] The tunnel's longitudinal segmented power supply and distribution topology is constructed as follows: the ultra-long tunnel is divided into N power supply sections according to its length, and each section is equipped with an independent intelligent power distribution unit. Adjacent power distribution units are connected by anti-electromagnetic interference composite cables, which employ a multi-layer shielding structure, and their core wires are made of graphene-modified copper-based composite material. Cable manufacturing process: copper powder and graphene oxide are ball-milled at a mass ratio of 98:2 for 4 hours at a speed of 300 rpm. Plasma sintering is then performed under argon protection (purity ≥99.99%) at a temperature of 800℃, a holding time of 30 minutes, and a pressure of 10 MPa. The sintered preform is then processed into wire using an equal channel angle extrusion (ECAP) process at a temperature of 400℃ and an extrusion ratio of 8:1.

[0046] Real-time multi-source data acquisition: A distributed fiber optic sensor network is deployed to acquire real-time energy consumption data of lighting, ventilation, and monitoring equipment in each power supply section, while simultaneously collecting tunnel traffic flow and ambient temperature and humidity data. The distributed fiber optic sensor network adopts Brillouin optical time-domain reflectometry (BOTDA) technology, with a sensor node spacing of ≤50 meters. Each sensor node is embedded in the tunnel sidewall, and the encapsulation shell is made of carbon nanotube-reinforced epoxy resin composite material.

[0047] Sensor network configuration: The fiber optic sensor network consists of distributed temperature sensors (DTS), distributed strain sensors (DAS), and distributed vibration sensors (DVS), with a sampling frequency of 1kHz.

[0048] The data acquisition cycle is 10 seconds, and the data is transmitted to the cloud management platform via 4G / 5G wireless network.

[0049] Dynamic energy consumption balance optimization: Based on an improved quantum genetic algorithm, energy consumption data is modeled, and the optimization objective is to minimize total network loss and load imbalance. The parameters of transformer taps and reactive power compensation devices in each power supply section are dynamically adjusted. The improved quantum genetic algorithm includes the use of qubit encoding to represent the transformer tap position and reactive power compensation amount.

[0050] Algorithm parameter settings: qubit encoding length is 10, population size is 50, and the number of iterations is 100. An adaptive rotation angle mechanism is introduced, with an adaptive rotation angle step size Δθ_t = θ_0 × e^{-γt}, where θ_0 = 0.2π and γ = 0.03. The catastrophe operator is triggered when the population diversity falls below 0.2, resetting 30% of the individual quantum states.

[0051] Fault self-healing control: When the power supply to a certain section is abnormal, a fast switching switch based on graphene composite conductor is activated to transfer the load to an adjacent section.

[0052] Switch design: The switch uses a graphene-carbon nanotube composite electrode with a contact resistance ≤10μΩ and a response time ≤50ms. The fault detection module is triggered by current surges (>150A) and voltage drops (<80% of rated value), and the switching logic is executed by the PLC controller.

[0053] The preparation of graphene composite conductors includes the following steps:

[0054] (1) The copper foil was immersed in a 0.5 mol / L silver nitrate solution to carry out a displacement reaction, and a layer of silver nanoparticles was generated on the surface;

[0055] (2) Methane was cracked at 550°C by chemical vapor deposition, and a vertical graphene array was grown on the silver nanoparticle layer.

[0056] (3) Gradient annealing under a protective atmosphere: heat up to 300℃ at 10℃ / min and hold for 1 hour, then heat up to 450℃ at 5℃ / min and hold for 2 hours.

[0057] The beneficial effects of this embodiment are as follows:

[0058] The conductivity of graphene composite cables is improved by more than 15% compared to traditional copper cables, and the electromagnetic interference suppression capability is improved by 20%. An improved quantum genetic algorithm, through an adaptive rotation angle mechanism and a catastrophe operator, increases the optimization convergence speed by 30% and reduces network loss by 12% (compared to 3.1% for traditional genetic algorithms). The response time of the graphene composite conductor fast-switching switch is ≤50ms, 5 times faster than traditional mechanical switches, and the power supply reliability is improved to 99.99%.

[0059] Example 2

[0060] Please see Figure 1 Based on Example 1, this example provides a technical solution: a method for optimizing the segmented power supply and distribution topology and dynamically balancing energy consumption in ultra-long tunnels, combining load prediction and self-diagnosis system energy consumption balancing, with the following implementation steps:

[0061] Capacity design: Based on the power distribution capacity formula for the k-th power supply section. Calculate the design capacity of each section. For example, the length of a certain section... =1.2 km, predicted peak traffic flow =800 vehicles / hour =1.5, =0.3, then =1.5×1.2×(1+0.3×ln(800))≈4.2MVA.

[0062] Dynamic correction factor Applications: The system dynamically adjusts based on historical traffic flow data, with a correction range of ±0.05, and uses an LSTM neural network to predict traffic flow changes in the next 15 minutes (error rate ≤8%).

[0063] Load prediction: The model embeds an LSTM neural network module, and the input layer contains feature vectors such as date type (weekday / weekend), time period (peak / off-peak), and weather status (sunny / rainy / foggy).

[0064] Model training and deployment: The training dataset consists of traffic flow and energy consumption data from the past three years. The model structure is a 3-layer LSTM unit (64 nodes per layer), with ReLU as the activation function and mean squared error (MSE) as the loss function. The prediction results are updated every 15 minutes, dynamically adjusting the transformer tap position and reactive power compensation.

[0065] Self-diagnostic system: Deploy a self-diagnostic system in the intelligent power distribution unit to diagnose winding loosening faults by monitoring the energy change rate in the 2-5kHz frequency band of the transformer vibration spectrum. When the change rate exceeds 15%, an early warning signal is triggered.

[0066] Early warning mechanism: The vibration sensor sampling frequency is 10kHz, the spectrum analysis adopts short time Fourier transform (STFT), and the early warning signal is transmitted to the operation and maintenance platform through the Modbus protocol.

[0067] Reactive power compensation: The reactive power compensation device adopts a modular multilevel converter structure, and its power module capacitor is a barium titanate@titanium dioxide core-shell structure dielectric material.

[0068] Capacitor material preparation: Tetrabutyl titanate was hydrolyzed to generate TiO2 sol, with the pH value controlled at 4.5 and the hydrolysis temperature at 80℃. A 5-10 nm thick TiO2 layer was coated onto the surface of BaTiO3 nanoparticles via electrostatic self-assembly, and the sintering temperature was 1350℃, with a nitrogen atmosphere for 2 hours.

[0069] The beneficial effects of this embodiment are as follows:

[0070] The dynamic correction factor β makes the capacity design more closely match the actual traffic flow demand, reducing equipment redundancy costs by 10%. The LSTM model, combined with multi-dimensional feature vectors, achieves a load prediction error rate of ≤8% (compared to 15% for traditional methods), and the dynamic adjustment strategy better meets actual needs. The self-diagnostic system can issue an early warning 30 minutes before a transformer winding loosening fault occurs, reducing the risk of sudden downtime and extending equipment life by 20%.

[0071] Example 3

[0072] Please see Figure 1 Based on Embodiment 1 or 2, this embodiment provides a technical solution: a method for segmented power supply and distribution topology optimization and dynamic energy consumption balancing in ultra-long tunnels, which also includes deploying a wind-solar hybrid power generation device at the tunnel entrance, integrating wind-solar hybrid power generation with high-efficiency heat dissipation. Implementation steps:

[0073] Wind-solar hybrid power generation: The wind-solar hybrid device deployed at the tunnel entrance is connected to the power supply and distribution system through a SiC-MOSFET DC-DC converter.

[0074] Equipment configuration: The wind turbine is a horizontal axis three-blade structure with a rated power of 5kW, a cut-in wind speed of 3m / s, and a cut-out wind speed of 25m / s.

[0075] The photovoltaic module is a monocrystalline silicon photovoltaic panel with a conversion efficiency of 22% and a maximum power point tracking (MPPT) efficiency of ≥98%.

[0076] Energy distribution: The wind-solar hybrid system prioritizes powering lighting and ventilation equipment, with surplus energy stored in lithium battery banks along the tunnel.

[0077] Energy storage system design: The lithium battery pack has a capacity of 100kWh, a cycle life of ≥5000 cycles, and a depth of discharge (DOD) of 80%.

[0078] The Energy Management System (EMS) uses a fuzzy control algorithm to dynamically adjust the power distribution between wind and solar power generation and energy storage systems.

[0079] Thermal management optimization: Diamond-aluminum composite metal material is used as the heat dissipation substrate for SiC-MOSFET, with a thermal conductivity ≥600 W / (m·K).

[0080] Heat dissipation design: The heat dissipation substrate is 3mm thick and coated with nano-level silver paste (thermal conductivity ≥10 W / (m·K)) to directly contact the SiC-MOSFET chip.

[0081] The heat dissipation system transfers heat to the outside of the tunnel through a liquid cooling cycle (the coolant is a mixture of deionized water and ethylene glycol), reducing the temperature rise by 20%.

[0082] The beneficial effects of this embodiment are as follows:

[0083] The wind-solar hybrid power system can generate up to 120 MWh / km per year, and combined with lithium battery energy storage, it increases the proportion of renewable energy to 40%. The heat dissipation performance of the diamond-aluminum composite material reduces the junction temperature of SiC-MOSFETs by 30°C and doubles the device lifespan. This reduces reliance on fossil fuels and reduces carbon dioxide emissions by approximately 800 tons per year (based on traditional power supply methods).

[0084] Comparative Example 1

[0085] Comparison of traditional power supply and distribution topology with genetic algorithms:

[0086] Power supply and distribution optimization was performed using traditional copper cables and a conventional genetic algorithm, without the use of graphene composite materials or quantum genetic algorithms.

[0087] Fault recovery relies on mechanical switches, with a response time of ≥300ms.

[0088] Technical drawbacks: Traditional copper cables have insufficient shielding performance, resulting in decreased signal transmission quality and a 10% increase in communication error rate. Ordinary genetic algorithms have slow convergence speed (requiring 150 iterations) and high network loss (3.5%), which is 12% higher than the solution of this invention. Mechanical switches have a response time ≥300ms and power supply reliability of only 99.5%, which is 0.49% lower than that of this invention.

[0089] Comparative Example 2

[0090] Comparison of power supply and distribution systems with no-load prediction and self-diagnostic systems:

[0091] The system did not employ an LSTM load prediction module or a transformer self-diagnostic system, relying solely on fixed capacity design and manual inspection.

[0092] Technical drawbacks: The fixed capacity design increases transformer load imbalance by 20%, resulting in significant energy waste (15% higher than this invention). Manual inspections take 7 days, and the average detection time for transformer winding loosening faults is delayed by more than 1 hour, posing a safety hazard.

Claims

1. A method for topology optimization and dynamic energy consumption balancing in segmented power supply and distribution systems for ultra-long tunnels, characterized in that, Includes the following steps: S1. Construct a longitudinal segmented power supply and distribution topology for the tunnel: Divide the tunnel into N power supply sections based on the tunnel length, deploy an independent intelligent power distribution unit in each section, and connect adjacent power distribution units through anti-electromagnetic interference composite cables; S2. Real-time acquisition of multi-source data: Real-time energy consumption data of lighting equipment, ventilation equipment and monitoring equipment in each power supply section are acquired through a distributed optical fiber sensor network, while tunnel traffic flow and ambient temperature and humidity data are also collected. S3. Dynamic energy consumption balance optimization: Based on the improved quantum genetic algorithm, energy consumption data is modeled, and the parameters of transformer taps and reactive power compensation devices in each power supply section are dynamically adjusted with the goal of minimizing total network loss and load imbalance. S4. Fault self-healing control: When the power supply of a certain section is abnormal, a fast switching switch based on graphene composite conductor is activated to automatically transfer the load of the faulty section to the power supply circuit of the adjacent section.

2. The method for segmented power supply and distribution topology optimization and dynamic energy consumption balancing in ultra-long tunnels according to claim 1, characterized in that, The electromagnetic interference-resistant composite cable adopts a multi-layer shielding structure, and its core wire is made of graphene-modified copper-based composite material. The preparation process includes: ball milling and mixing copper powder and graphene oxide at a mass ratio of 98:2, forming a preform by plasma sintering under argon protection, and then processing it into wire by equal channel angle extrusion process.

3. The method for topology optimization and dynamic energy consumption balancing of segmented power supply and distribution in ultra-long tunnels according to claim 1, characterized in that, In the segmented power supply and distribution topology, the power distribution capacity design of the k-th power supply segment satisfies the formula: in For design capacity, For segment length, To predict peak traffic volume, As the baseline capacity factor, This is a dynamic correction factor.

4. The method for topology optimization and dynamic energy consumption balancing of segmented power supply and distribution in ultra-long tunnels according to claim 1, characterized in that, The distributed optical fiber sensor network uses Brillouin optical time-domain reflectometry, with sensor nodes embedded in the tunnel sidewall at a spacing of no more than 50 meters, and its encapsulation shell is made of carbon nanotube-reinforced epoxy resin composite material.

5. The method for topology optimization and dynamic energy consumption balancing of segmented power supply and distribution in ultra-long tunnels according to claim 1, characterized in that, The improved quantum genetic algorithm includes: The transformer tap position and reactive power compensation are represented by quantum bit encoding; Introducing an adaptive rotation angle mechanism, the decay formula for the rotation angle step size Δθ at iteration number t is: Δθ_t = θ_0 × e^{-γt} Where θ_0 is the initial rotation angle, ranging from 0.1π to 0.3π; γ is the decay coefficient, ranging from 0.02 to 0.05; and t is the current iteration number. Add a catastrophe operator to reset 30% of the individual quantum states when the population diversity falls below a threshold.

6. The method for segmented power supply and distribution topology optimization and dynamic energy consumption balancing in ultra-long tunnels according to claim 1, characterized in that, The preparation of the graphene composite conductor includes the following steps: S41. A copper foil is immersed in a 0.5 mol / L silver nitrate solution to carry out a displacement reaction, thereby generating a layer of silver nanoparticles on the surface; S42. Vertical graphene arrays were grown on silver nanoparticle layers by chemical vapor deposition to crack methane at 550 °C. S43. Perform gradient annealing under a protective atmosphere: heat to 300℃ at 10℃ / min and hold for 1 hour, then heat to 450℃ at 5℃ / min and hold for 2 hours.

7. The method for topology optimization and dynamic energy balance of segmented power supply and distribution in ultra-long tunnels according to claim 1, characterized in that, Also includes: In step S3, a load prediction module is embedded, which uses an LSTM neural network to predict load changes in the next 15 minutes based on historical traffic flow data. The input layer of the prediction model includes feature vectors for date type, time period, and weather status.

8. The method for topology optimization and dynamic energy balance of segmented power supply and distribution in ultra-long tunnels according to claim 1, characterized in that, Each intelligent power distribution unit has a built-in self-diagnostic system that diagnoses winding loosening faults by monitoring the energy change rate in the 2-5kHz frequency band of the transformer vibration spectrum. When the change rate exceeds 15%, an early warning signal is triggered.

9. The method for topology optimization and dynamic energy balance of segmented power supply and distribution in ultra-long tunnels according to claim 1, characterized in that, The reactive power compensation device adopts a modular multilevel converter structure, and its power module capacitor uses a barium titanate@titanium dioxide core-shell structure dielectric material. The preparation method includes: Tetrabutyl titanate is hydrolyzed to generate TiO2 sol; A 5-10 nm thick TiO2 layer was coated onto the surface of BaTiO3 nanoparticles via electrostatic self-assembly. Dense ceramics are formed by sintering in a nitrogen atmosphere at 1350℃.

10. The method for topology optimization and dynamic energy consumption balancing of segmented power supply and distribution in ultra-long tunnels according to claim 1, characterized in that, A wind-solar hybrid power generation device is deployed at the tunnel entrance. Its output power is connected to the power supply and distribution system through a SiC-MOSFET-based DC-DC converter. The heat dissipation substrate of the converter is made of diamond-aluminum composite metal material with a thermal conductivity ≥600W / (m·K).

Images