Control method of regenerative braking energy utilization system for double-current system circuit

Abstract

The invention provides a control method of a regenerative braking energy utilization system for a double-current system line, which belongs to the technical field of electrified railway traction power supply and comprises the following steps of: constructing a hardware system structure comprising two pieces of power fusion equipment; and establishing a game model with the lowest system electric charge as a target, and optimizing to obtain peak clipping set values of alternating current and direct current traction. Voltage and current data are collected in real time, and filtering processing is performed by using a sliding time window method. And calculating the instantaneous power of the AC and DC loads according to the filtered data. And establishing a fuzzy neural network model, and determining system working modes including energy storage charging, power transfer and energy storage discharging according to the load power and the peak clipping set value. And establishing an adaptive power distribution equation set, and solving to obtain power reference values of the converters and the energy storage system. Power outer loop control is adopted, a corresponding current reference value is calculated, a self-adaptive current tracker is used for generating a switching signal, and utilization control over regenerative braking energy is achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of electrified railway traction power supply, and in particular, relates to a control method for a regenerative braking energy utilization system for a dual-current line. Background Art

[0002] The AC / DC dual-current urban rail transit traction power supply mode has achieved seamless connection between the conventional urban rail transit lines and the urban rail transit express lines, realizing "zero-distance" passenger transfers, greatly improving the rail transit service level, and is the future development direction of urban rail transit. However, electric locomotive braking often uses regenerative braking. In urban rail transit, due to the short station spacing, frequent train starts and braking, the braking energy is considerable, and at the same time, in the AC traction power supply system, high-power regenerative braking energy is also generated. However, when the locomotives on the same arm cannot absorb the regenerative braking energy, the excess regenerative braking energy will be injected into the power grid through the traction substation or dissipated by the braking resistor, resulting in energy waste. And due to the existence of the neutral section and the electrical segmentation, the energy in the dual-current system cannot flow to each other, resulting in a low regenerative braking energy utilization rate.

[0003] Energy storage technology can realize demand-side power management, peak shaving and valley filling, and load smoothing in the field of rail transit; however, the regenerative braking energy utilization system for the dual-current traction power supply system is still blank, and due to the natural energy barrier between DC and AC lines, the energy cannot be transmitted directly, and the power flow cannot be managed uniformly.

[0004] Therefore, how to efficiently utilize the regenerative braking energy in the dual-current traction power supply system has become one of the key technologies to improve the energy efficiency of rail transit systems.

[0005] At present, the utilization schemes of regenerative braking energy mainly include the following:

[0006] 1. DC bus type. In the DC traction power supply system, the feedback current of the regenerative motor can be directly connected to the DC bus to realize the utilization of regenerative energy on the DC side. However, this solution can only utilize the regenerative power on the DC side, but not the regenerative power on the AC side.

[0007] 2. Bidirectional converter type. In the AC traction power supply system, a bidirectional converter is installed between the AC power supply arm and the traction load to achieve bidirectional transmission of regenerative power between AC and DC. However, this solution requires the addition of a large-capacity bidirectional converter, which is costly.

[0008] 3. Energy storage type. In the AC / DC hybrid power supply system, an energy storage system is installed between the AC / DC power supply arms to absorb regenerative power and release it when needed. This solution can take into account the utilization of regenerative energy on both the AC and DC sides, but it requires additional large-capacity energy storage devices, and the system complexity is relatively high.

[0009] Although the above-mentioned solutions have achieved the utilization of regenerative braking energy to a certain extent, there are still the following main problems:

[0010] 1. Insufficient utilization of regenerative power on the AC and DC sides. Existing solutions can only utilize regenerative power on the DC side, or require the addition of large-capacity bidirectional converters or energy storage devices, which results in high system costs.

[0011] 2. Lack of intelligent power distribution and coordinated control. Most existing solutions use manually set power distribution strategies, which are unable to perceive and adaptively adjust the system operation status in real time, making it difficult to achieve optimal energy utilization.

[0012] 3. The control strategy is relatively simple. Most existing solutions use relatively simple control strategies, such as current loop control, power loop control, etc., which are difficult to meet the complex requirements of the regenerative braking energy utilization system.

[0013] Therefore, it is urgent to develop a new type of regenerative braking energy utilization system that can make full use of the regenerative electric energy on the AC and DC sides, and adopt intelligent power distribution and coordinated control strategies to improve the energy utilization efficiency and economic performance of the system. Summary of the invention

[0014] In view of this, the present invention provides a control method for a regenerative braking energy utilization system for a dual-current circuit, which can solve the technical problem that the prior art is difficult to achieve efficient utilization and intelligent management of regenerative electric energy on the AC and DC sides.

[0015] The present invention is achieved in that:

[0016] The present invention provides a control method for a regenerative braking energy utilization system for a dual-current circuit, comprising the following steps:

[0017] S10, constructing a hardware structure of a regenerative braking energy utilization system for a dual-current system circuit, wherein the system includes a power integration device I and a power integration device II;

[0018] S20, using a particle swarm optimization algorithm to establish a game model with the lowest system electricity fee as the objective function, and obtaining the peak shaving set value of the AC traction station and the peak shaving set value of the DC traction station through iterative optimization calculation;

[0019] S30, collecting voltage and current data of the left and right power supply arms and the DC power supply arm of the AC traction station, and establishing a real-time database;

[0020] S40, filtering the voltage and current data in the real-time database using a sliding time window method to obtain filtered voltage and current data;

[0021] S50, calculating the instantaneous power of the AC power supply arm load and the instantaneous power of the DC power supply arm load according to the filtered voltage and current data;

[0022] S60, establishing a fuzzy neural network model, inputting the instantaneous power of the AC load, the instantaneous power of the DC load and the peak clipping setting value into the fuzzy neural network model, and outputting a determined system operating mode;

[0023] S70, according to the system working mode, establish an adaptive power allocation equation group, and solve to obtain power reference values ​​of the α-side AC-DC converter, the β-side AC-DC converter, the left-side AC-DC converter, the right-side DC-DC converter, and the energy storage system;

[0024] S80, adopting power outer loop control, establishing a current reference value calculation model based on the power reference value, and calculating and obtaining current reference values ​​of the α-side AC-DC converter, the β-side AC-DC converter, the left-side AC-DC converter, the right-side DC-DC converter, and the energy storage system;

[0025] S90, comparing the current reference value with the actual current value of the corresponding port, using an adaptive current tracker to obtain a modulation signal, and converting the modulation signal into a switching signal of each converter through space vector pulse width modulation technology to achieve utilization control of regenerative braking energy.

[0026] On the basis of the above technical solution, the control method of the regenerative braking energy utilization system for the dual-current circuit of the present invention can also be improved as follows:

[0027] Among them, the working modes include energy storage charging mode, power transfer mode, and energy storage discharging mode.

[0028] The objective function of the game model is specifically expressed as follows:

[0029]

[0030] Where, J is the total system cost; T is the optimization period, in hours; C grid，t is the power purchase cost of the power grid at time t; C dem，t is the demand electricity cost at time t; C loss，t is the system loss cost at time t.

[0031] in:

[0032] Cgrid，t =P buy，t P buy,t -p sell，t P sell，t ;

[0033] In the formula, p buy，t is the power purchase price of the power grid at time t, in yuan / kWh; P buy,t is the purchased power at time t, in kW; p sell，t is the on-grid electricity price at time t, in yuan / kWh; P sell，t is the electricity sales power at time t, in kW.

[0034] C dem，t =p dem max{P load，t};

[0035] In the formula, p dem is the demand electricity price, in yuan / kW; P load，t is the load power at time t, in kW.

[0036] C loss,t =p buy，t (P loss,AC，t +P loss,DC,t +P loss，conv，t );

[0037] Where P loss,AC，t is the power loss on the AC side at time t; P loss,DC，t is the power loss on the DC side at time t; P loss，conv，t is the power loss of the converter at time t.

[0038] Constraints:

[0039] 1. Power balance constraints:

[0040] P buy,t -P sell，t +P ESS,t =P load，t +P loss,t ;

[0041] Where P ESS，t is the charging and discharging power of the energy storage system at time t, charging is positive and discharging is negative.

[0042] 2. Energy storage system constraints:

[0043] E ESS,min ≤E ESS，t ≤E ESS,max ;

[0044] P ESS,min ≤P ESS,t ≤PESS,max ;

[0045] In the formula, E ESS，t is the remaining capacity of the energy storage system at time t; E ESS,min and E ESS,max are the lower and upper limits of the energy storage system capacity respectively; P ESS,min and P ESS,max are the lower and upper limits of the energy storage system power respectively.

[0046] The power integration device I is composed of an α-side step-down transformer, a β-side step-down transformer, an α-side AC-DC converter, and a β-side AC-DC converter.

[0047] Furthermore, the AC port of the α-side AC-DC converter is connected to the secondary side of the α-side step-down transformer, the primary side of the α-side step-down transformer is connected to the left power supply arm of the AC traction station, the DC port of the α-side AC-DC converter is connected to the DC port of the β-side AC-DC converter, the AC port of the β-side AC-DC converter is connected to the secondary side of the β-side step-down transformer, and the primary side of the β-side step-down transformer is connected to the right power supply arm of the AC traction station.

[0048] Furthermore, the power integration device II is composed of a left-side step-down transformer, a left-side AC-DC converter, a right-side DC-DC converter and an energy storage system.

[0049] Furthermore, the AC port of the left AC-DC converter is connected to the secondary side of the left step-down transformer, the primary side of the left step-down transformer is connected to the left AC power supply arm of the neutral section, the DC port of the left AC-DC converter is connected to the capacitor port of the right DC-DC converter, and the inductor port of the right DC-DC converter is connected to the right DC power supply arm.

[0050] Furthermore, the energy storage system includes a DC-DC converter and a supercapacitor, the inductor end of the DC-DC converter is connected to the supercapacitor, and the capacitor end is connected to the DC port of the left AC-DC converter and the capacitor end of the right DC-DC converter.

[0051] Furthermore, the adaptive power distribution equation group includes an AC power distribution equation, a DC power distribution equation, an energy storage power distribution equation, a power balance equation, and a power regulation equation; the AC power distribution equation is used to calculate the power distribution of the AC side converter, the input is the AC load power and the peak clipping set value, and the output is the power reference value of the α-side and β-side converters, which is specifically expressed as follows:

[0052]

[0053] P β，ref =PAC，total -P α，ref ;

[0054] Where P AC，total is the total AC load power; P AC，high is the AC peak clipping set value; k1, k2 are adaptive coefficients, which are estimated online by the least squares method; ε1 is the error compensation term.

[0055] The DC power distribution equation is used to calculate the power distribution of the DC side converter. The input is the DC load power and the peak clipping setting value, and the output is the power reference value of the DC converter. It is specifically expressed as follows:

[0056]

[0057] Where P DC，total is the total DC load power; P DC,high is the DC peak clipping setting value; k3, k4 are adaptive coefficients; jωL is the inductive reactance compensation item; ε2 is the error compensation item.

[0058] The energy storage power allocation equation is used to calculate the power allocation of the energy storage system. The input is the total power deviation of the system, and the output is the power reference value of the energy storage system. It is specifically expressed as follows:

[0059] P ESS,ref =k5(P AC，total +P DC，total -P AC，high -P DC,high )+k6∫(P AC，total +P DC，total )dt+ε3;

[0060] Where k5, k6 are adaptive coefficients; ε3 is the error compensation term.

[0061] The power balance equation is used to ensure system power balance, and is specifically expressed as follows:

[0062] P α，ref +P β，ref +P DC，ref +P ESS，ref =P AC，total +P DC，total .

[0063] The power regulation equation is used to realize dynamic regulation of power, which is specifically expressed as follows:

[0064]

[0065] Where P ref is the power reference value; P actual is the actual power; k7, k8 are the adjustment coefficients; ε4 is the error compensation term.

[0066] Furthermore, the AC power distribution equation is used to calculate the power distribution of the AC side converter, the input is the AC load power and the peak shaving setting value, and the output is the power reference value of the α-side and β-side converters; the DC power distribution equation is used to calculate the power distribution of the DC side converter, the input is the DC load power and the peak shaving setting value, and the output is the power reference value of the DC converter; the energy storage power distribution equation is used to calculate the power distribution of the energy storage system, the input is the system total power deviation, and the output is the power reference value of the energy storage system; the power balance equation is used to ensure the system power balance; the power regulation equation is used to realize dynamic regulation of power.

[0067] Furthermore, the specific structure of the fuzzy neural network model is a five-layer feedforward network, including an input layer, a fuzzy layer, a rule layer, a defuzzification layer and an output layer.

[0068] Optionally, it also includes a fuzzy adjustment equation group module and a fuzzy attention equation group module. The fuzzy adjustment equation group module includes a membership function adjustment equation, a rule weight adjustment equation and an output scale factor adjustment equation;

[0069] The membership function adjustment equation is used to dynamically adjust the membership function. The input is the system error and the error change rate. The output is the adjusted membership function parameters. It is specifically expressed as follows:

[0070]

[0071] In the formula, μ ij (x) is the jth membership function of the ith input variable; c ij is the central value; ij is the width parameter.

[0072] The rule weight adjustment equation is used to dynamically adjust the weight of the fuzzy rule. The input is the system performance index, and the output is the adjusted rule weight. It is specifically expressed as follows:

[0073]

[0074] In the formula, w k is the weight of the kth rule; η is the learning rate; ΔJ is the change of the performance index; is the sensitivity of the output to the weight.

[0075] The output scale factor adjustment equation is used to dynamically adjust the output scale factor. The input is the system output deviation, and the output is the adjusted scale factor. It is specifically expressed as follows:

[0076]

[0077] In the formula, Ko is the output scale factor; α, β are adjustment parameters; e is the system output deviation.

[0078] The fuzzy attention equation group module includes an attention weight calculation equation, a context vector calculation equation and an output fusion equation;

[0079] The attention weight calculation equation is used to calculate the importance weights of different input features. The input is the system state vector and the output is the attention weight. It is specifically expressed as follows:

[0080]

[0081] In the formula, α i is the attention weight of the i-th feature; s t-1 is the system status at the previous moment; h i is the i-th input feature; score() is the relevance scoring function.

[0082] The context vector calculation equation is used to generate a context vector taking into account the attention mechanism, which is specifically expressed as follows:

[0083]

[0084] In the formula, c t is the context vector at time t.

[0085] The output fusion equation is used to fuse the context vector with the current state, which is specifically expressed as follows:

[0086] o t =tanh(W c [c t ;s t ]);

[0087] In the formula, o t is the output after fusion; W c is the weight matrix; [c t ;s t ] represents the concatenation of the context vector and the current state.

[0088] The method of converting the modulation signal into a switching signal of each converter by using the space vector pulse width modulation technology specifically includes the following steps:

[0089] 1. Convert the three-phase modulated signal into a space vector:

[0090]

[0091] 2. Determine the sector where the space vector is located:

[0092]

[0093] In the formula, v α , v β is the component of the space vector in the α β coordinate system.

[0094] 3. Calculate the action time of the basic vector:

[0095]

[0096] T0=T s -T1-T2;

[0097] Where, T s is the switching period; θ is the angle between the space vector and the starting edge of the sector.

[0098] 4. Generate PWM waveform: Generate the conduction signal of each switch tube according to the action time sequence.

[0099] Compared with the prior art, the control method of the regenerative braking energy utilization system for the dual-flow circuit provided by the present invention has the following beneficial effects:

[0100] 1. Efficient utilization of regenerative power on the AC and DC sides is achieved. By building a hardware structure including two sets of power integration equipment and an energy storage system, bidirectional conversion and storage of regenerative power on the AC and DC sides can be achieved, maximizing the utilization rate of regenerative power.

[0101] 2. Adopt intelligent power distribution and coordinated control strategy. The present invention adopts particle swarm optimization algorithm to determine the optimal AC and DC peak clipping setting value, establishes fuzzy neural network model to adaptively judge the system working mode, and designs adaptive power distribution equations to achieve intelligent distribution and coordinated control of renewable electric energy.

[0102] 3. The control strategy is more advanced and efficient. The present invention adopts adaptive current tracking control at the current loop node and uses space vector pulse width modulation technology to generate switching signals. Compared with traditional current loop control and power loop control, it has stronger dynamic response capability and higher control accuracy.

[0103] 4. Improvement of system energy efficiency and economy. Through the above technical measures, the present invention can maximize the use of regenerative braking energy and reduce the energy loss of the system, thereby significantly reducing the operating electricity cost of the system and improving the overall energy efficiency and economic performance of the system.

[0104] In summary, the regenerative braking energy utilization system control method proposed in the present invention has been innovatively designed in terms of hardware topology, power allocation strategy, control algorithm, etc., to achieve efficient utilization and intelligent management of regenerative electric energy, and solve the technical problem that the existing technology is difficult to achieve efficient utilization and intelligent management of regenerative electric energy on the AC and DC sides. BRIEF DESCRIPTION OF THE DRAWINGS

[0105] Figure 1 A flow chart of the method provided by the present invention;

[0106] Figure 2 A schematic diagram of the topological structure of a regenerative braking energy utilization system for a dual-flow circuit in an embodiment;

[0107] Figure 3 The topological structure of the power fusion device 1 in the embodiment;

[0108] Figure 4 It is the topological structure of the power fusion device II in the embodiment;

[0109] Figure 5 It is a timing diagram of system power distribution in the embodiment;

[0110] Figure 6 It is a system power allocation stacking diagram in the embodiment;

[0111] Figure 7 It is a block diagram of a dual-loop control strategy of an AC-DC converter with an H-bridge structure, in which a voltage outer loop predicts a current inner loop;

[0112] Figure 8 PI control strategy block diagram of a half-bridge DC-DC converter in an embodiment. DETAILED DESCRIPTION

[0113] In order to make the purpose, technical solution and advantages of the embodiments of the present invention more clear, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0114] like Figure 1 As shown, it is a flow chart of a control method of a regenerative braking energy utilization system for a dual-flow circuit provided by the present invention, and the method comprises the following steps:

[0115] S10, constructing a hardware structure of a regenerative braking energy utilization system for a dual-current system circuit, the system comprising a power integration device I and a power integration device II;

[0116] S20, using a particle swarm optimization algorithm to establish a game model with the lowest system electricity fee as the objective function, and obtaining the peak shaving set value of the AC traction station and the peak shaving set value of the DC traction station through iterative optimization calculation;

[0117] S30, collecting voltage and current data of the left and right power supply arms and the DC power supply arm of the AC traction station, and establishing a real-time database;

[0118] S40, filtering the voltage and current data in the real-time database using a sliding time window method to obtain filtered voltage and current data;

[0119] S50, calculating the instantaneous power of the AC power supply arm load and the DC power supply arm load based on the filtered voltage and current data;

[0120] S60, establishing a fuzzy neural network model, inputting the AC load instantaneous power, the DC load instantaneous power and the peak shaving setting value into the fuzzy neural network model, and outputting a determined system operating mode;

[0121] S70, according to the system working mode, establish an adaptive power allocation equation group, and solve to obtain power reference values ​​of the α-side AC-DC converter, the β-side AC-DC converter, the left-side AC-DC converter, the right-side DC-DC converter, and the energy storage system;

[0122] S80, using power outer loop control, establishing a current reference value calculation model based on the power reference value, and calculating and obtaining current reference values ​​of the α-side AC-DC converter, the β-side AC-DC converter, the left-side AC-DC converter, the right-side DC-DC converter, and the energy storage system;

[0123] S90, comparing the current reference value with the actual current value of the corresponding port, using an adaptive current tracker to obtain a modulation signal, and converting the modulation signal into a switching signal of each converter through space vector pulse width modulation technology to achieve utilization control of regenerative braking energy.

[0124] The specific implementation methods of the above steps are described in detail below:

[0125] Step S10: Building the system hardware structure. The purpose of this step is to establish the hardware environment of the regenerative braking energy utilization system.

[0126] First, construct the power integration device I. The power integration device I consists of an α-side step-down transformer, a β-side step-down transformer, an α-side AC-DC converter, and a β-side AC-DC converter. Among them, the AC port of the α-side AC-DC converter is connected to the secondary side of the α-side step-down transformer, and the primary side of the α-side step-down transformer is connected to the left power supply arm of the AC traction station. The DC port of the α-side AC-DC converter is connected to the DC port of the β-side AC-DC converter. The AC port of the β-side AC-DC converter is connected to the secondary side of the β-side step-down transformer, and the primary side of the β-side step-down transformer is connected to the right power supply arm of the AC traction station. This topological structure realizes power integration between the left and right power supply arms of the AC traction station.

[0127] Next, construct the power integration device II. The power integration device II consists of a step-down transformer on the left, an AC-DC converter on the left, a DC-DC converter on the right, and an energy storage system. The AC port of the AC-DC converter on the left is connected to the secondary side of the step-down transformer on the left, and the primary side of the step-down transformer on the left is connected to the left AC power supply arm of the neutral section. The DC port of the AC-DC converter on the left is connected to the capacitor port of the DC-DC converter on the right, and the inductor port of the DC-DC converter on the right is connected to the DC power supply arm on the right.

[0128] Finally, the energy storage system is constructed. The energy storage system includes a DC-DC converter and a supercapacitor, where the inductor end of the DC-DC converter is connected to the supercapacitor, and the capacitor end is connected to the DC port of the left AC-DC converter and the capacitor end of the right DC-DC converter. Such an energy storage system can store regenerative braking energy and release it when needed.

[0129] In general, by constructing the above hardware structure, a hardware environment including two sets of power integration equipment and an energy storage system is formed, which lays the foundation for the subsequent regenerative braking energy utilization control.

[0130] Step S20: using a particle swarm optimization algorithm to establish a game model with the lowest system electricity fee as the objective function.

[0131] The purpose of this step is to use the particle swarm optimization algorithm to find the peak shaving setting value that minimizes the total electricity cost of the system. First, a game model with the total electricity cost of the system as the objective function is established. The total electricity cost of the system includes three parts: the power purchase cost of the power grid C grid,t 、Demand electricity cost C dem，t and system loss cost C loss,t . It can be expressed as:

[0132]

[0133] Where T is the optimization period in hours.

[0134] Power purchase cost from the power grid C grid，t It can be expressed as:

[0135] C grid,t =p buy，t P buy,t -p sell，t P sell，t ;

[0136] In the formula, p buy，t is the power purchase price of the power grid at time t, in yuan / kWh; P buy，t is the purchased power at time t, in kW; p sell，t is the on-grid electricity price at time t, in yuan / kWh; P sell，tis the electricity sales power at time t, in kW.

[0137] Demand electricity cost C dem，t It can be expressed as:

[0138] C dem，t =p dem max{P load，t};

[0139] In the formula, p dem is the demand electricity price, in yuan / kW; P load，t is the load power at time t, in kW.

[0140] System loss cost C loss，t It can be expressed as:

[0141] C loss,t =p buy，t (P loss,AC，t +P loss,DC，t +P loss，conv，t );

[0142] Where P loss,AC，t is the power loss on the AC side at time t; P loss,DC，t is the power loss on the DC side at time t; P loss，conv，t is the power loss of the converter at time t.

[0143] Next, the particle swarm optimization algorithm is used to iteratively optimize the objective function to obtain the AC peak-shaving setting value P that minimizes the total electricity cost. AC，high and DC peak clipping setting value P DC,high .

[0144] The particle swarm optimization algorithm is a swarm intelligence algorithm that simulates the foraging behavior of bird flocks to find the optimal solution. In the algorithm, each particle represents a potential solution. The particle continuously updates its position and velocity based on its own historical experience and the experience of the entire group, and finally converges to the global optimal solution. In this step, each particle represents a group of P AC，high and P DC,high ,The algorithm updates the position and velocity of the particles through continuous iteration, and ,ultimately finds the peak shaving setting value that minimizes the total electricity cost of the ,system.

[0145] Step S30: Collect voltage and current data of the left and right power supply arms and the DC power supply arm of the AC traction station and establish a real-time database.

[0146] The purpose of this step is to obtain real-time voltage and current data during system operation to provide a basis for subsequent power calculation and control. Specifically, it includes the following:

[0147] First, collect the voltage and current data of the left and right power supply arms of the AC traction station. Use voltage and current sensors to collect the voltage v of the left and right power supply arms of the AC traction station in real time. a (t), v b (t), v c (t) and current i a (t),i b (t),i c (t).

[0148] Secondly, collect the voltage and current data of the DC power supply arm. Use the voltage and current sensor to collect the voltage v of the DC power supply arm in real time. DC and current i DC .

[0149] Finally, the collected voltage and current data are recorded in real time in the database to provide a basis for subsequent data processing and power calculation.

[0150] Through this step, a database containing the real-time voltage and current data of the AC and DC power supply arms was established, laying the foundation for real-time monitoring and control of the system.

[0151] Step S40: filtering the voltage and current data in the real-time database using a sliding time window method to obtain filtered voltage and current data.

[0152] The purpose of this step is to filter the original voltage and current data collected in step S30 to remove high-frequency interference and obtain smoother and more stable data. The specific method is as follows:

[0153] First, set the filtering time window length N. According to the dynamic characteristics of the system, select a suitable filtering time window length, for example, N=100, that is, 1 second.

[0154] Then, the collected voltage and current data are subjected to sliding average filtering. i , use the data in the time window centered on t to calculate the average value as the filtered data at that moment The formula is as follows:

[0155]

[0156] In the formula, is the filtered data at time t, x i is the original data at time i, and N is the length of the time window.

[0157] Finally, the filtered voltage and current data are stored in a real-time database for use in subsequent steps.

[0158] Through this step, the high-frequency interference in the original voltage and current data is removed, and smoother and more stable data is obtained, which provides a good input for subsequent power calculation and control.

[0159] Step S50: Calculate the instantaneous power of the AC power supply arm load and the DC power supply arm load based on the filtered voltage and current data.

[0160] The purpose of this step is to calculate the instantaneous power of the AC and DC power supply arms to provide a basis for subsequent power distribution and control. The specific calculation method is as follows:

[0161] First, calculate the instantaneous power P of the AC power supply arms on both sides AC，total The AC power supply arm voltage v obtained in step S40 a (t), v b (t), v c (t) and current i a (t),i b (t),i C (t) filtered data, using the formula:

[0162]

[0163] Calculate the total instantaneous power P of the AC power supply arms on both sides AC，total Where, T is the power calculation time window.

[0164] Secondly, calculate the instantaneous power P of the DC power supply arm DC，total The DC power supply arm voltage v obtained in step S40 DC and current i DC Filter data using the formula:

[0165] P DC，total =v DC i DC ;

[0166] Calculate the instantaneous power P of the DC power supply arm DC，total .

[0167] Through this step, the instantaneous power data of the AC and DC power supply arms are obtained, which provides an important basis for subsequent power distribution and control.

[0168] Step S60: Establish a fuzzy neural network model to convert the instantaneous power P of the AC load into AC，total , the instantaneous power P of the DC load DC，total And the peak clipping setting value P AC，high and P DC,high The fuzzy neural network model is input and a determined system operating mode is output.

[0169] The purpose of this step is to establish a fuzzy neural network model to automatically determine the current working mode of the system based on the AC and DC load power and peak shaving set values. The specific method is as follows:

[0170] First, a five-layer feedforward fuzzy neural network model is constructed. The model includes input layer, fuzzy layer, rule layer, defuzzification layer and output layer. The input layer receives P AC，total , P DC，total , P AC，high and P DC,high The output layer gives the current system operating mode, including energy storage charging mode, power transfer mode and energy storage discharging mode.

[0171] Secondly, establish the fuzzy membership function. Use Gaussian fuzzy membership function for input variables:

[0172]

[0173] In the formula, μ ij (x) is the jth membership function of the ith input variable; c ij is the central value; ij is the width parameter. Adaptive adjustment of the membership function is achieved by dynamically adjusting these two parameters.

[0174] Next, build a fuzzy rule base. Design fuzzy rules based on engineering experience and assign a certain weight W to each rule. k By dynamically adjusting rule weights:

[0175]

[0176] Among them, η is the learning rate, ΔJ is the change of performance index, The fuzzy rule base is optimized by outputting sensitivity to weights.

[0177] Finally, the fuzzy reasoning mechanism and central focus method are used for defuzzification to obtain the final system working mode.

[0178] Through this step, an intelligent fuzzy neural network model is established, which can automatically determine the current system working mode according to the real-time AC and DC load power and peak shaving set values, providing a basis for subsequent power allocation and control.

[0179] Step S70: According to the system working mode, an adaptive power allocation equation group is established, and the equation group is solved to obtain power reference values ​​of the α-side AC-DC converter, the β-side AC-DC converter, the left-side AC-DC converter, the right-side DC-DC converter, and the energy storage system.

[0180] The purpose of this step is to establish a set of adaptive power allocation equations according to the system operating mode determined in step S60, and calculate the power reference values ​​of each converter and energy storage system. The specific method is as follows:

[0181] First, establish the AC power distribution equation. The AC power distribution equation is used to calculate the power distribution of the AC-DC converter on the α side and the β side. The input is the total AC load power P AC，total and AC peak clipping setting value P AC，high , the output is the power reference value P of the α-side and β-side converters α，ref and P β，ref Specifically:

[0182]

[0183] P β，ref =P AC，total -P a，ref ;

[0184] Where k1, k2 are adaptive coefficients, and ε1 is the error compensation term.

[0185] Secondly, establish the DC power distribution equation. The DC power distribution equation is used to calculate the power distribution of the left AC-DC converter and the right DC-DC converter. The input is the total DC load power P DC，total and DC peak clipping setting value P DC,high , the output is the power reference value P of the DC converter DC,ref Specifically:

[0186]

[0187] Where k3, k4 are adaptive coefficients, jωL is the inductive reactance compensation term, and ε2 is the error compensation term.

[0188] Next, establish the energy storage power allocation equation. The energy storage power allocation equation is used to calculate the power allocation of the energy storage system. The input is the total power deviation of AC and DC, and the output is the power reference value P of the energy storage system. ESS，ref Specifically:

[0189] P ESS,ref =k5(P AC，total +P DC，total -P AC，high -P DC,high )+k6∫(P AC，total +P DC，total )dt+ε3;

[0190] Where k5, k6 are adaptive coefficients, and ε3 is the error compensation term.

[0191] Next, establish the power balance equation. The power balance equation is used to ensure system power balance:

[0192] P α，ref +P β，ref +P DC，ref +P ESS，ref =P AC，total +P DC，total ;

[0193] Finally, establish the power regulation equation. The power regulation equation is used to achieve dynamic power regulation:

[0194]

[0195] Where P ref is the power reference value, P actual is the actual power, k7 and k8 are the adjustment coefficients, and ε4 is the error compensation term.

[0196] By solving this set of adaptive power allocation equations, the power reference values ​​of each converter and energy storage system can be calculated to meet the power balance and dynamic characteristics requirements of the system.

[0197] Step S80: adopting power outer loop control, establishing a current reference value calculation model based on the power reference value, and calculating the current reference values ​​of the α-side AC-DC converter, the β-side AC-DC converter, the left-side AC-DC converter, the right-side DC-DC converter and the energy storage system.

[0198] The purpose of this step is to establish a current reference value calculation model based on the power reference values ​​of each converter and energy storage system calculated in step S70, and obtain the current reference value of each port to provide a basis for subsequent current closed-loop control. The specific method is as follows:

[0199] First, a power outer loop control structure is established. With the power outer loop control, the power reference value calculated in step S70 is used as the outer loop control target, and a corresponding control loop is established.

[0200] Secondly, a current reference value calculation model is established. The current reference value is calculated based on the power reference value of each port and the voltage data. Taking the α-side AC-DC converter as an example, its current reference value i α，ref It can be calculated as:

[0201]

[0202] Where P α，ref is the α-side converter power reference value calculated in step S70, v α is the input voltage of the α-side converter.

[0203] Finally, the current reference values ​​of the α-side AC-DC converter, the β-side AC-DC converter, the left-side AC-DC converter, the right-side DC-DC converter and the energy storage system are calculated respectively.

[0204] Through this step, a current reference value calculation model based on the power outer loop is established, which provides the required current reference value for subsequent current tracking control.

[0205] Step S90: Compare the current reference value with the actual current value of the corresponding port, use an adaptive current tracker to obtain a modulation signal, and convert the modulation signal into a switching signal of each converter through space vector pulse width modulation technology to achieve utilization control of regenerative braking energy.

[0206] The purpose of this step is to achieve accurate current control of each converter to ensure that the regenerative braking energy can be effectively utilized. The specific method is as follows:

[0207] First, an adaptive current tracker is constructed. The current reference value calculated in step S80 is compared with the actual current value, and a modulation signal is obtained by using an adaptive PI controller. The parameters of the adaptive PI controller can be dynamically adjusted according to the system operating conditions to improve the current tracking performance.

[0208] Secondly, the space vector pulse width modulation technology is used to generate the switching signal. The specific steps are as follows:

[0209] 1. Convert the three-phase modulated signal into a space vector representation:

[0210]

[0211] 2. Determine the sector where the space vector is located:

[0212]

[0213] In the formula, v α , v β is the component of the space vector in the αβ coordinate system.

[0214] 3. Calculate the action time of the basic vector:

[0215]

[0216] T0=T s -T1-T2;

[0217] Where, T s is the switching period, and θ is the angle between the space vector and the starting edge of the sector.

[0218] 4. Generate PWM waveform: Generate the conduction signal of each switch tube according to the calculated action time sequence.

[0219] Finally, the output switching signal drives the α-side AC-DC converter, the β-side AC-DC converter, the left-side AC-DC converter, the right-side DC-DC converter and each switch tube of the energy storage system to realize the utilization control of regenerative braking energy.

[0220] Through this step, adaptive current tracking control and space vector pulse width modulation technology are used to accurately control the switching state of each converter, ensuring that the regenerative braking energy can be effectively converted and stored between the AC and DC sides, thereby improving energy utilization efficiency.

[0221] Specifically, the principle of the present invention is:

[0222] 1. Principle of hardware structure design. The present invention constructs a hardware structure including two sets of power integration equipment and an energy storage system. Power integration equipment I realizes power integration between the left and right power supply arms on the AC side through the α-side and β-side AC-DC converters; power integration equipment II realizes power conversion between the AC and DC sides through the left AC-DC converter and the right DC-DC converter; the energy storage system can absorb and release regenerative braking energy. Such a hardware structure can maximize the conversion channel of regenerative electric energy between the AC and DC sides and improve the utilization efficiency of regenerative electric energy.

[0223] 2. Principle of power allocation strategy. The present invention adopts particle swarm optimization algorithm to determine the optimal peak shaving setting value on the AC and DC sides, and designs an adaptive power allocation equation group. The AC power allocation equation calculates the power allocation of the α-side and β-side converters according to the total AC load and the peak shaving setting value; the DC power allocation equation calculates the power allocation of the DC converter according to the total DC load and the peak shaving setting value; the energy storage power allocation equation calculates the power allocation of the energy storage system according to the total system power deviation. This adaptive power allocation strategy can make full use of renewable electric energy and reduce the total electricity cost of the system.

[0224] 3. Principle of control algorithm. The present invention adopts fuzzy neural network model to realize intelligent judgment of system working mode, establishes current reference value calculation model based on power outer loop control, and adopts adaptive current tracking control and space vector pulse width modulation technology to realize precise control of each converter. Fuzzy neural network can adaptively judge the current working mode of the system according to AC and DC load power and peak clipping setting value; power outer loop control can ensure the dynamic balance of power of each port; adaptive current tracking control and space vector pulse width modulation technology can realize precise control of converter switch state, and improve the dynamic response speed and stability of renewable electric energy utilization.

[0225] In order to better understand and implement the present invention, an embodiment of a specific application scenario of the present invention is provided below: a certain city subway system adopts AC / DC hybrid power supply mode, the total length of the line is 50 kilometers, and the average passenger capacity is 2,000 people / train. The subway train adopts regenerative braking technology, which will generate a large amount of regenerative electric energy during the braking process. In order to make full use of this regenerative electric energy, the city subway company decided to adopt the regenerative braking energy utilization system control method proposed in the present invention.

[0226] First, if Figure 2-4 As shown, the hardware structure of the regenerative braking energy utilization system is as follows:

[0227] The power integration equipment I consists of a step-down transformer on the α side, a step-down transformer on the β side, an AC-DC converter on the α side, and an AC-DC converter on the β side. The primary side of the step-down transformer on the α side is connected to the left power supply arm of the AC traction station, and the primary side of the step-down transformer on the β side is connected to the right power supply arm of the AC traction station. The DC ports of the two AC-DC converters are connected to achieve power integration between the left and right power supply arms on the AC side.

[0228] The power integration device II consists of a step-down transformer on the left, an AC-DC converter on the left, a DC-DC converter on the right, and an energy storage system. The primary side of the step-down transformer on the left is connected to the AC power supply arm on the left side of the neutral section, and the inductor end of the DC-DC converter on the right side is connected to the DC power supply arm on the right side. This enables bidirectional power conversion between the AC and DC sides.

[0229] The energy storage system includes a 2MW, 500kWh supercapacitor bank connected to the left AC-DC converter and the right DC-DC converter via a DC-DC converter. The energy storage system can absorb the electrical energy generated by regenerative braking and release it when needed.

[0230] The α-side AC-DC converter, the β-side AC-DC converter and the left-side AC-DC converter adopt an H-bridge structure, and the right-side DC-DC converter and the DC / DC converter of the energy storage system adopt a half-bridge bidirectional DC / DC converter topology.

[0231] The parameters of the system are set as follows:

[0232] The voltage level of the AC power supply system is AC25kV and the frequency is 50Hz; the voltage level of the DC power supply system is DC750V. The rated power of each converter is 2MW, and the switching frequency is 1kHz. The resistance of the power supply arms on the left and right sides of the AC traction station is 0.2Ω and 0.25Ω respectively, and the resistance of the DC power supply arm is 0.15Ω. The power purchase price of the power grid is p buy =0.6 yuan / kWh, on-grid electricity price p sell =0.4 yuan / kWh, demand price pdem =100 yuan / kW.

[0233] Next, the control flow of the system is as follows:

[0234] 1. Use particle swarm optimization algorithm to determine the optimal AC and DC peak shaving setpoints. The objective function of the algorithm is to minimize the total electricity cost of the system, and the constraints include power balance and energy storage system capacity limitations. After 50 iterations of optimization, the AC peak shaving setpoint P is finally obtained. AC，high =1.5MW, DC peak clipping setting value P DC,high =1MW.

[0235] 2. Collect the voltage and current data of the left and right power supply arms and the DC power supply arm of the AC traction station, and filter them using the sliding time window method. The time window length N = 100, i.e. 1 second. After filtering, the total AC load power P is obtained. AC，total and the total DC load power P DC，total The following is an analysis of the data from 17:00 to 18:00 on a certain day, as shown in Table 1: Table 1 Data collected within 1 hour

[0236] Time (h) <![CDATA[P AC，total (MW)]]> <![CDATA[P DC,total (MW)]]> 17:00 1.8 1.2 17:10 1.9 1.1 17:20 2.0 1.0 17:30 2.1 0.9 17:40 2.0 1.1 17:50 1.9 1.2 18:00 1.8 1.3

[0237] Figure 5 (System power distribution time series diagram) shows the changing trends of AC load power, DC load power and energy storage system power during 17:00-18:00. The red line represents the AC load power, the blue line represents the DC load power, and the green dotted line represents the energy storage system power. This diagram clearly shows the relationship and changing rules of the three powers, especially the charging and discharging status of the energy storage system at different times. The green shaded area in the figure intuitively shows the working status of the energy storage system.

[0238] 3. Establish a fuzzy neural network model to judge the system working mode. The model includes a five-layer feedforward structure. The input layer receives P AC，total , P DC，total , P AC，high and P DC,high , the output layer gives the current system working mode. For the data within the hour of 17:00-18:00 mentioned above, the model determines that:

[0239] 17:00-17:20, the system is in energy storage charging mode, charging regenerative braking energy into the energy storage system;

[0240] 17:30-17:50, the system is in power transfer mode, transferring the regenerative power on the AC side to the DC side;

[0241] At 18:00, the system is in energy storage discharge mode, releasing the electrical energy in the energy storage system to the DC power supply arm.

[0242] 4. According to the system working mode, establish the adaptive power distribution equation group. The AC power distribution equation is:

[0243]

[0244] P β，ref =P AC，total -P α，ref ;

[0245] The DC power distribution equation is:

[0246]

[0247] The energy storage power allocation equation is:

[0248] P ESS,ref =1.8(P AC，total +P DC，total -1.5-1.0)+0.6∫(P AC，total +P DC，total )dt+0.04;

[0249] Among them, k1 = 1.2, k2 = 0.5, k3 = 1.5, k4 = 0.4, k5 = 1.8, k6 = 0.6 are adaptive coefficients. These equations can dynamically calculate the power reference value of each port according to the real-time AC and DC load characteristics to achieve the optimal distribution of regenerative power.

[0250] 5. Based on the power reference value, establish a current reference value calculation model. Taking the α-side AC-DC converter as an example, its current reference value is:

[0251]

[0252] Among them, v α is the input voltage of the α-side converter. The current reference values ​​of other converters and energy storage systems can be calculated similarly.

[0253] 6. Adaptive current tracking control and space vector pulse width modulation technology are used to achieve precise control. The proportional coefficient and integral coefficient of the adaptive PI controller are dynamically adjusted according to the real-time error, which can quickly eliminate the current tracking error. Space vector pulse width modulation technology can accurately generate the switching signal of each converter to achieve efficient conversion and storage of regenerated electric energy between the AC and DC sides.

[0254] In addition, you can Figure 7 and Figure 8As shown, the AC-DC converter adopts a dual-loop control of a power outer loop and a current inner loop, and then obtains a switching signal for controlling the AC-DC converter through PWM modulation; the DC-DC converter adopts PI control, and the reference current value is compared with the actual current value through PI control, and then through a hysteresis comparator to obtain a switching signal for controlling the DC-DC converter.

[0255] Through the above control process, the operation of the system in the hour from 17:00 to 18:00 is as follows:

[0256] 17:00-17:20, the system is in energy storage charging mode. Output power P of the α-side converter αref =0.6MW, β-side converter output power P β，ref =1.2MW, energy storage system absorbed power P ESS,ref =-0.8MW. In this stage, the electric energy generated by regenerative braking is mainly absorbed by the energy storage system.

[0257] 17:30-17:50, the system is in power transfer mode. α，ref =1.2MW, P β，ref =0.9MW, P DC，ref =0.8MW, P ESS,ref = 0. In this stage, the regenerative electric energy on the AC side is converted to the DC side through the α-side and β-side converters, and part of the DC load is met.

[0258] At 18:00, the system is in energy storage and discharge mode. α，ref =1MW, P β，ref =0.8MW, P DC，ref =1MW,P ESS,ref =0.5MW. In this stage, the energy storage system releases electric energy to the DC side to meet the needs of the DC load.

[0259] Figure 6 (System power distribution stacked diagram) The power distribution of the α-side converter, β-side converter, and DC converter is shown in the form of a stacked area diagram. The pink area represents the α-side converter power, the blue area represents the β-side converter power, and the light green area represents the DC converter power. This stacked display method can intuitively show the composition of the total power and the relative proportion changes of each part.

[0260] Through the above power distribution and precise control, the operation status of the system in the hour from 17:00 to 18:00 is shown in Table 2 below:

[0261] Table 2 Operation status table

[0262]

[0263] As can be seen from the table above, through the control method of the regenerative braking energy utilization system proposed by the present invention, the total electricity cost of the system is only 6750 yuan in the hour from 17:00 to 18:00, which is 16.7% lower than the 8100 yuan when no measures are taken. Among them, the electricity purchase cost is reduced from 7200 yuan to 6600 yuan, the electricity sales income is increased from 1100 yuan to 1350 yuan, and the system loss cost is reduced from 800 yuan to 800 yuan. It can be seen that the control method can not only improve the utilization rate of regenerative electric energy, but also significantly reduce the system operation cost, and has good economic benefits.

[0264] The following is an embodiment 2 of the present invention. The main difference between embodiment 2 and embodiment 1 is that the process in embodiment 2 is a simpler process, and the working mode and working condition are determined by calculation. The hardware structure of the regenerative braking energy utilization system for the dual-flow circuit adopted by both is the same. In order to more conveniently describe embodiment 2, the meaning of the variables in this embodiment 2 is independent of the above. The steps of this embodiment 2 specifically include:

[0265] Step S10: Setting the peak clipping setting value of the AC traction and DC traction peak clipping setting value

[0266] Step S20: Collect the voltage and current of the AC and DC substations respectively, and calculate the instantaneous power P of the load on the power supply arms on both sides of the AC T-α and P T-β , and the instantaneous power P of the DC power supply arm load T-DC , according to the instantaneous power P of the AC load T-α and P T-β , and the peak shaving setting value range, determine the working mode of the system, the working mode includes energy storage charging mode, power transfer mode, and energy storage discharging mode;

[0267] Step S30: Calculate the power reference values ​​of the α-side AC-DC converter, the β-side AC-DC converter, the left-side AC-DC converter, the right-side DC-DC converter, and the energy storage system according to the working mode and specific working conditions. Know

[0268] Step S40: Obtain the current reference values ​​of the α-side AC-DC converter, the β-side AC-DC converter, the left-side AC-DC converter, the right-side DC-DC converter and the energy storage system through the power outer loop and

[0269] Step S50: According to each current reference value and the actual current value of the corresponding port, the corresponding modulation signal is obtained by using the current inner loop; and each modulation signal is converted into a switching signal of the α-side AC-DC converter, the β-side AC-DC converter, the left-side AC-DC converter, the right-side DC-DC converter and the energy storage system through PWM modulation.

[0270] In step S10, the peak clipping setting value of the AC traction and DC traction peak clipping setting value The calculation method uses an intelligent algorithm to obtain the optimal value. The specific process is as follows:

[0271] A. Establish the objective function of minimizing the system electricity cost

[0272] minf0=C grid +C dem ;

[0273]

[0274] C dem =[max(P ACdem,t )+max(P DCdem,t )]·c dem / N day ;

[0275] C grid The sum of the electricity charges of the traction substation; C dem The sum of the electricity charges required for traction substation; P ACgridbuy，t P is the electric power obtained by the AC traction substation from the power grid; DCgridbuy，t P is the electric power obtained by the DC traction substation from the power grid; ACgridfed，t P is the electric energy power fed back to the power grid from the AC traction substation; ACdem，t P is the power required for AC traction substation; DCdem，t is the power required for DC traction substation; C buy，t Indicates the unit price of power obtained from the three-phase power grid, c fed，t represents the unit price of power returned to the grid through the AC traction substation; c dem Indicates the unit electricity price of peak demand power, N day Indicates the number of days per month that the system runs, for example, 30 days.

[0276] B. Taking the converter capacity, energy storage capacity, system power balance, and AC traction peak shaving set value and DC traction peak shaving set value as constraints, the intelligent algorithm is used for iterative optimization to obtain the AC traction peak shaving set value and DC traction peak shaving set value under the condition of the lowest system electricity cost.

[0277] In step S20, the method for determining the working mode and working condition of the energy router is:

[0278] A: If P T-α +P T-β +P T-DC <0, the system is judged to be in energy storage charging mode, which is judged as working condition 1, and the power reference value of each converter is:

[0279]

[0280] B: If The energy router is judged to be in power transfer mode. Since the power of AC traction is obviously greater than that of DC traction, P will not appear in power transfer mode. T-α +P T-β <0;

[0281] b1) If P T-α +P T-β <0, It is determined to be working condition 2, and the power reference value of each converter is:

[0282]

[0283] b2) If P T-α +P T-β <0, It is determined to be working condition 3, and the power reference value of each converter is:

[0284]

[0285] b3) If P T-DC <0, it is judged as working condition 4, and the power reference value of each converter is:

[0286]

[0287] b4) If It is determined to be working condition 5, and the power reference value of each converter is:

[0288]

[0289] b5) If It is determined to be working condition 6, and the power reference value of each converter is:

[0290]

[0291] b6) If P T-DC <0, it is judged as working condition 7, and the power reference value of each converter is:

[0292]

[0293] b7) If It is determined to be working condition 8, and the power reference value of each converter is:

[0294] C: If It is determined that the system is in energy storage and discharge mode, which is working condition 9. The power reference value of each converter is:

[0295]

[0296] In step S40, the calculation formula of each current reference value is:

[0297]

[0298] Among them, U α , U β , U left , U DC and U ESS It is the voltage of the α-side AC-DC converter, the β-side AC-DC converter, the left-side AC-DC converter, the right-side DC-DC converter and the energy storage system.

[0299] like Figure 7 and Figure 8 As shown, in step S50, the AC-DC converter adopts a power outer loop predicting current inner loop dual-loop control, and then obtains a switching signal for controlling the AC-DC converter through PWM modulation; the DC-DC converter adopts PI control, based on the reference current value The actual current value i x Compared with the PI control, the switching signal of the DC-DC converter is obtained through the hysteresis comparator.

[0300] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A control method for a regenerative braking energy utilization system for a dual-flow circuit, characterized in that: The following steps are involved: S10, constructing a hardware structure of a regenerative braking energy utilization system for a dual-current system circuit, wherein the system includes a power integration device I and a power integration device II; S20, using a particle swarm optimization algorithm to establish a game model with the lowest system electricity fee as the objective function, and obtaining the peak shaving set value of the AC traction station and the peak shaving set value of the DC traction station through iterative optimization calculation; S30, collecting voltage and current data of the left and right power supply arms and the DC power supply arm of the AC traction station, and establishing a real-time database; S40, filtering the voltage and current data in the real-time database using a sliding time window method to obtain filtered voltage and current data; S50, calculating the instantaneous power of the AC power supply arm load and the instantaneous power of the DC power supply arm load according to the filtered voltage and current data; S60, establishing a fuzzy neural network model, inputting the instantaneous power of the AC load, the instantaneous power of the DC load and the peak clipping setting value into the fuzzy neural network model, and outputting a determined system operating mode; S70, according to the system working mode, establish an adaptive power allocation equation group, and solve to obtain power reference values ​​of the α-side AC-DC converter, the β-side AC-DC converter, the left-side AC-DC converter, the right-side DC-DC converter, and the energy storage system; S80, adopting power outer loop control, establishing a current reference value calculation model based on the power reference value, and calculating and obtaining current reference values ​​of the α-side AC-DC converter, the β-side AC-DC converter, the left-side AC-DC converter, the right-side DC-DC converter, and the energy storage system; S90, comparing the current reference value with the actual current value of the corresponding port, using an adaptive current tracker to obtain a modulation signal, and converting the modulation signal into a switching signal of each converter through space vector pulse width modulation technology to achieve utilization control of regenerative braking energy.

2. The control method of a regenerative braking energy utilization system for a dual-current circuit according to claim 1, characterized in that: The working modes include energy storage charging mode, power transfer mode, and energy storage discharging mode.

3. The control method of a regenerative braking energy utilization system for a dual-current circuit according to claim 1, characterized in that: The power integration device I is composed of an α-side step-down transformer, a β-side step-down transformer, an α-side AC-DC converter, and a β-side AC-DC converter.

4. The control method of a regenerative braking energy utilization system for a dual-current circuit according to claim 3 is characterized in that: The AC port of the α-side AC-DC converter is connected to the secondary side of the α-side step-down transformer, the primary side of the α-side step-down transformer is connected to the left power supply arm of the AC traction station, the DC port of the α-side AC-DC converter is connected to the DC port of the β-side AC-DC converter, the AC port of the β-side AC-DC converter is connected to the secondary side of the β-side step-down transformer, and the primary side of the β-side step-down transformer is connected to the right power supply arm of the AC traction station.

5. The control method of a regenerative braking energy utilization system for a dual-current circuit according to claim 4 is characterized in that: The power integration device II is composed of a step-down transformer on the left, an AC-DC converter on the left, a DC-DC converter on the right, and an energy storage system.

6. The control method of the regenerative braking energy utilization system for dual-current circuit according to claim 5 is characterized in that: The AC port of the left AC-DC converter is connected to the secondary side of the left step-down transformer, the primary side of the left step-down transformer is connected to the left AC power supply arm of the neutral section, the DC port of the left AC-DC converter is connected to the capacitor port of the right DC-DC converter, and the inductor port of the right DC-DC converter is connected to the right DC power supply arm.

7. The control method of a regenerative braking energy utilization system for a dual-current circuit according to claim 6, characterized in that: The energy storage system includes a DC-DC converter and a supercapacitor, wherein the inductor end of the DC-DC converter is connected to the supercapacitor, and the capacitor end is connected to the DC port of the left AC-DC converter and the capacitor end of the right DC-DC converter.

8. The control method of the regenerative braking energy utilization system for dual-current circuit according to claim 7 is characterized in that: The adaptive power distribution equation group includes an AC power distribution equation, a DC power distribution equation, an energy storage power distribution equation, a power balance equation and a power regulation equation; wherein the AC power distribution equation is used to calculate the power distribution of the AC side converter, the input is the AC load power and the peak shaving setting value, and the output is the power reference value of the α-side and β-side converters; the DC power distribution equation is used to calculate the power distribution of the DC side converter, the input is the DC load power and the peak shaving setting value, and the output is the power reference value of the DC converter; the energy storage power distribution equation is used to calculate the power distribution of the energy storage system, the input is the system total power deviation, and the output is the power reference value of the energy storage system; the power balance equation is used to ensure the power balance of the system; the power regulation equation is used to realize dynamic regulation of power.

9. The control method of the regenerative braking energy utilization system for dual-current circuit according to claim 8, characterized in that: The AC power distribution equation is used to calculate the power distribution of the AC side converter, the input is the AC load power and the peak shaving setting value, and the output is the power reference value of the α-side and β-side converters; the DC power distribution equation is used to calculate the power distribution of the DC side converter, the input is the DC load power and the peak shaving setting value, and the output is the power reference value of the DC converter; the energy storage power distribution equation is used to calculate the power distribution of the energy storage system, the input is the system total power deviation, and the output is the power reference value of the energy storage system; the power balance equation is used to ensure the power balance of the system; the power regulation equation is used to realize dynamic regulation of power.

10. The control method of the regenerative braking energy utilization system for dual-current circuit according to claim 9, characterized in that: The specific structure of the fuzzy neural network model is a five-layer feedforward network, including an input layer, a fuzzy layer, a rule layer, a defuzzification layer and an output layer.

Images