CN112066594A - An orderly control method for large-scale electric heat pump based on μPMU device - Google Patents

Abstract

The invention discloses a method for controlling orderly power consumption of a large-scale electric heat pump based on a μPMU device, which utilizes the fast and accurate measurement data of the μPMU, selects characteristic quantities in combination with an electric heat pump model and start-up characteristics, and formulates a characteristic judgment of the start-up state of the electric heat pump. Then, according to the power constraint coupled by the distribution network voltage constraint model considering the cluster electric heat pump, the orderly power consumption control method is carried out for the electric heat pump load group under the three working states. The method controls the electric heat pump to stop in an orderly manner, reduces the steady-state power and controls the starting sequence, and avoids the superposition of the electric heat pump starting shock and exceeding the limit.

Description

An orderly control method for large-scale electric heat pump based on μPMU device

technical field

The invention relates to a method for controlling electricity consumption of an electric heat pump, in particular to a method for controlling electricity consumption in an orderly manner for a large-scale electric heat pump based on a μPMU device.

Background technique

Electric heat pump is a heat pump system that drives heat pump mechanical operation by electric energy. It has the advantages of wide application range, low operating cost, stable performance and environmental protection. At present, it has been widely used for heating and hot water supply in housing. The characteristics of centralized power consumption by users have a significant impact on the daily peak-to-valley difference and seasonal peak-to-valley difference, especially in the heating season when the load increases sharply, coupled with the irregularity of residents' heating and travel, the peak load of the distribution network. The valley difference becomes larger and the load three-phase imbalance problem is obvious. In the low-voltage distribution network, the start-up shock caused by the short-term start-up process of the electric heat pump will cause obvious current inrush pulses and voltage-drop pulses.

Affected by the irregular operation of the electric heat pump and the start-up shock, the instantaneous fluctuation of the three-phase voltage can easily exceed the power quality regulations. A serious voltage drop will cause the start-up failure of the electric heat pump. The start-up shock not only affects the power quality of the low-voltage distribution network. It may make the electric heat pump itself unable to start, and may also affect the operation and start of other electric heat pumps and other electrical equipment. The existing technology guarantees electricity consumption by extensively constructing and transforming the distribution network, but this method is not only economical, but also has a long construction period and is not suitable for practical application.

SUMMARY OF THE INVENTION

Purpose of the invention: The present invention proposes a method for controlling the orderly power consumption of a large-scale electric heat pump based on μPMU equipment, which reduces the start-up impact when the electric heat pump is used on a large scale.

Technical scheme: the technical scheme adopted in the present invention is a large-scale electric heat pump based on μPMU device orderly electricity control method, comprising the following steps:

1) Modeling of electric heat pump heating system;

2) Identify the state of the electric heat pump based on μPMU measurement;

3) Orderly electricity consumption strategy of electric heat pump based on measurement and identification.

Said step 1) in the electric heat pump heating system

The functional relationship between the electric power P _cmp of the compressor and the rotational speed N and the refrigerant flow rate m _r :

表示排气压力与吸气压力之比；In the above formula, c ₁ is the suction specific volume of the compressor, n is the variable index of the compression process, η _cmp is the ratio of the indicated power of the compressor to the power consumption,

Represents the ratio of exhaust pressure to suction pressure;

The heat transfer model of the condenser is as follows:

In the formula, C _{r, sh, cond} is the specific heat coefficient of the refrigerant; T ₂ is the temperature of the refrigerant flowing into the condenser; T _cond is the superheat temperature of the condenser; T ₃ is the temperature of the refrigerant flowing out of the condenser; r _cond is The structural radius of the plate condenser; A _i,cond is the heat exchange area of the plate condenser; _Tw is the outlet temperature of the hot water side of the condenser; U _sc,cond , U _sh,cond , U _tp,cond are the condensation is the comprehensive coefficient of heat transfer in cold water measurement, superheated zone and two-phase zone of the boiler, and _Tw is the temperature of hot water;

The heat exchange of the evaporator satisfies:

Q _eva =Q _tp +Q _sh =Q _a (3)

In the above formula, Q _eva is the heat transferred to the refrigerant in the evaporator, and Q _a is the heat absorbed by the evaporator from the heat source medium.

The complex variable model of the d-q coordinate system of the asynchronous motor is described as:

分别定子电压、定子电流、转 子电流、定子磁链和转子磁链的复数向量；R_s、L_s、M、R_r、L_r分别为定子 电阻、定子电感、定转子互感、转子电阻和转子电感；l_s＝L_s-M和l_r＝L_r-M分 别为定子和转子漏电感；ω_r为转子电角速度，单位为rad/s；j是复数单位。In the above formula,

Complex vectors of stator voltage, stator current, rotor current, stator flux linkage and rotor flux linkage, respectively; R _s , L _s , M, R _r , L _r are stator resistance, stator inductance, stator-rotor mutual inductance, rotor resistance and rotor, respectively Inductance; l _s =L _s -M and l _r =L _r -M are the stator and rotor leakage inductances, respectively; ω _r is the rotor electrical angular velocity, in rad/s; j is a complex unit.

In the step 1), the closed interval functional relationship of the active power P _s with respect to the time t in the starting process of the electric heat pump is:

P _s (t)=-1160(t-0.2) ² +P _max t∈[t ₀ ,t ₁ ] (5)

In the above formula, t ₀ is the initial start time, t ₁ is the time when the electric heat pump enters a steady state after starting up, and P _max is the maximum starting power of the electric heat pump.

In the step 2), the constraint of the voltage U _i of the head-end node is first established, and the voltage U _i of the head-end node of the station area needs to satisfy:

U _i ≥{max(U _lim.k )|T _ek (ω _maxrk ,U _lim.k )=T _m (ω _maxrk ),k∈[1,n]} (6)

In the above formula, U _lim.k is the critical value when the electric heat pump load at node k can start normally, T _ek (ω _maxrk , U _lim.k ) is the electromagnetic torque acting on the motor rotor in the kth electric heat pump load, ω _maxrk is the maximum rotor speed when T _ek takes the maximum value, U _lim.k is the critical starting voltage at node k, and T _m (ω _maxrk ) is the mechanical torque of the electric heat pump at the per-unit value ω _maxrk of the asynchronous motor speed.

Then judge the start-up state of the electric heat pump according to the power shock and voltage sag. The power shock criterion is as follows:

In the above formula, P(t) is the load power at the current time t; P(t-1) is the load power at the previous measurement time; N is the measurement frequency at time t, and the power change takes n measurement data in turn Differences and summations are obtained; P'(t) is the rate of change of the load power at the current time t; λ ₁ and λ ₂ are the thresholds corresponding to the criteria.

The voltage sag criterion is:

In the formula, λ ₃ and λ ₄ are the thresholds used to describe the proximity of U _t-1 and U _t-2 , and N is the measurement frequency at time t.

In described step 2), utilize the quantity of quadratic function in electric heat pump start-up impact power, identify the number of electric heat pumps in start-up state.

In the step 3), the load voltage is converted into load power, as follows:

The voltage and injected active power of the head-end node n satisfy:

U _n =k _P ·[PP(t ₀ )]+U _n (t ₀ ) (9)

In the formula, U _n is the voltage required by the head-end node n; P is the injected power required by the head-end node _n , Un (t ₀ ), P (t ₀ ) are the current node voltage and power measured at time t ₀ , k _p is the slope, and when one of the quantities is known, the value of the other quantity can be calculated.

The steps of implementing the orderly power consumption strategy are as follows:

1) Obtain load information according to measurement and identification methods and determine load groups in different states;

2) Prioritize the electric heat pump loads of the two equipment groups in startup and running states;

3) Determine the order and quantity of controlled electric heat pump equipment.

In the described step 2), the superimposed peak value of the electric heat pump startup impulse power started at the same moment exceeds the limit, and the steady-state voltage priority, the different temperature priority and the current sequence position number are added to obtain the startup sequence number.

In the step 2), for the power exceeding the limit caused by the superimposed partial starting power when the load of the steady-state operation of the distribution transformer is relatively large, the current sequence position number, the priority of the electric heat pump sorted according to the average value of the applied voltage, and the priority of different temperatures Add up to get the closing sequence number.

Beneficial effects: The orderly power consumption control method of the present invention is based on the state identification of the electric heat pump and the fast measurement of the μPMU, and proposes an orderly power consumption control method of the electric heat pump that optimizes the start-up impact, using the fast and accurate measurement data of the μPMU, combined with the electric heat pump. Heat pump model and start-up characteristics Select the characteristic quantities and formulate the characteristic criteria of the start-up state of the electric heat pump. Then, according to the power constraints coupled by the distribution network voltage constraint model considering the cluster electric heat pump, the load groups of the electric heat pump under the three working states are evaluated. Sequence electricity control method. The method controls the electric heat pump to stop in an orderly manner, reduces the steady-state power and controls the starting sequence, and avoids the superposition of the electric heat pump starting shocks exceeding the limit.

Description of drawings

Fig. 1 is the heating cycle diagram of the electric heat pump unit;

Figure 2 is a schematic diagram of heat transfer in an evaporator;

Figure 3 is an equivalent circuit diagram of the T-type asynchronous motor;

Fig. 4 is the mechanical characteristic diagram of the asynchronous motor;

Figure 5 is an equivalent circuit diagram of a radial low-voltage distribution network;

Fig. 6 is the electric heat pump port equivalent circuit diagram of node k;

Figure 7 is a simplified Thevenin equivalent circuit diagram;

Figure 8 is a line equivalent diagram of a low-voltage distribution network;

Fig. 9 is the flow chart of the control method of orderly electricity consumption of electric heat pump;

Figure 10 is a graph of thermal behavior and power characteristics of the electric heat pump load.

Detailed ways

1. Modeling of electric heat pump heating system

The main internal components of the electric heat pump unit are evaporator, compressor, four-way valve, condenser, and expansion valve. Among them, the compressor, condenser and evaporator are the main heat exchange and energy consumption components that constitute the electric heat pump unit:

A. Compressor: It converts mechanical energy into internal energy of fluid by continuously rotating and doing work, which is called the heart of the electric heat pump unit. The compressor is the power source in the refrigerant cycle. The rotor of the compressor performs adiabatic compression of the refrigerant vapor from the evaporator and continuously pushes the vapor with higher pressure into the condenser.

B. Evaporator and condenser: It is an important heat exchange component in the electric heat pump unit. The evaporator is a process in which the refrigerant and the heat source medium convectively evaporate and absorb heat, so that the refrigerant obtains heat energy. The condenser is the part that the refrigerant releases heat to the cooling water through the process of condensation and heat release, and realizes the heat transfer of the refrigerant.

1 is an evaporator; 2 is a compressor; 3 is a four-way valve; 4 is a condenser; 5 is an expansion valve; 6 is a shut-off valve; 7 is an indoor radiator; T _a is the ambient temperature, _Tw is the hot water temperature; Q _o is the output heating capacity; T _eva,i , T _eva,o are the refrigerant temperatures flowing into and out of the evaporator; T _cond,i , T _cond,o are The temperature of the refrigerant flowing into and out of the condenser; Tw, _i and Tw, _o are the temperatures of the cooling water flowing into and out of the condenser; p _eva,i and p _eva,o are the pressures of the medium flowing into and out of the evaporator; p _cond,i and p _cond,o are the pressures of the medium flowing into and out of the condenser.

As shown in Figure 2, a partition model is generally used to describe the heat exchange process in the evaporator: it is divided into two-phase zone and superheat zone. The refrigerant in the evaporating tube absorbs heat in the two-phase region, and after absorbing the heat, the refrigerant boils and vaporizes into refrigerant vapor. Because the heat transfer coefficient of the refrigerant in the superheated zone is low, and the heat exchange per unit area of the superheated zone is also much smaller than that of the two-phase zone. Therefore, the heat source medium entering the evaporator mainly completes the heat exchange to the refrigerant in the two-phase region.

Heat exchange on the heat source medium side of the evaporator:

Q _a =ζ·C _p,a ·m _a ·(T _a -T ₀ ) (10)

In the formula, Ta is the temperature of the incoming heat source medium, ℃; ζ is the moisture separation coefficient; C _{p,a is} the specific heat capacity of the medium, which can be _1.003kJ /(kg*K); The mass of the heat source medium, and m _a =ρ·V; T ₀ is the temperature of the medium flowing out of the evaporator, in °C.

Heat exchange on the refrigerant side of the evaporator:

The heat transfer equation on the refrigerant side of the evaporator in the two-phase zone:

In the above formula, γ _i is the fouling coefficient in the tube; a _tp is the two-phase heat transfer coefficient of the refrigerant; D _in is the cross-sectional area of the heat exchange tube; T _W2 is the wall temperature of the two-phase zone; T _eva.i is the liquid entering the evaporator. Refrigerant temperature; l _tp is the length of the two-phase region.

The heat transfer equation of the refrigerant side in the superheated zone of the evaporator:

In the formula, a _V is the single-phase heat transfer coefficient of the refrigerant; T _W1 is the wall temperature of the superheated zone; T _eva.o is the refrigerant vapor temperature; l _sh is the length of the superheated zone.

It can be seen that the heat Q _a absorbed by the evaporator from the heat source medium is mainly determined by the temperature and flow rate of the heat source medium. Equations (11) to (12) show that the heat that the evaporator can transfer to the refrigerant is mainly determined by the structural parameters of the two-phase zone and the superheat zone of the evaporator.

Mathematical model of compressor

In order to ensure the normal operation of the refrigerant cycle, the compressor consumes electrical power and inputs mechanical energy to the refrigerant in adiabatic compression, and continuously pushes the refrigerant vapor formed in the evaporator to be sent to the condenser after adiabatic compression. The compressor drives the entire refrigerant cycle.

The relationship between the input power of the compressor and the pressure in the refrigerant cycle is as follows:

η_cmp为压缩机指示功率和耗电功率的比值，取为0.64；

表示排气压力与吸气压力之比，即压缩比；n为压缩过程多变指数，绝热压缩过 程中n为常数。In the above formula, P _cmp is the electric power consumed by the compressor, kW; the frequency f is 50HZ, s is the slip rate of the asynchronous motor; _{V d} is the theoretical gas delivery volume of the compressor (m ³ ); value is

η _cmp is the ratio of the indicated power of the compressor to the power consumption, which is taken as 0.64;

Represents the ratio of the exhaust pressure to the suction pressure, namely the compression ratio; n is the variable index of the compression process, and n is a constant in the adiabatic compression process.

The refrigerant mass flow m _r satisfies:

In the formula, N=60f(1-s) is the compressor rotational speed (rpm); c ₁ is the compressor suction specific volume (m ³ /kg).

From the above equations (13) and (14), the functional relationship of the compressor electric power P _cmp with respect to the rotational speed N and the refrigerant flow rate m _r can be established:

It can be seen that in the process of the compressor consuming electric energy and participating in the refrigerant cycle, its operating power and speed directly affect the flow rate of the refrigerant and the pressure (also internal energy) of the refrigerant vapor.

Mathematical model of condenser

The high-temperature and high-pressure refrigerant vapor after adiabatic compression by the compressor is pushed into the condenser, which is a heat exchange component that can release heat from the refrigerant to the cooling water to transfer the heat energy of the refrigerant. The condenser often uses a plate heat exchanger. According to the conservation of energy, the heat exchange model of the condenser is as follows:

In the formula, C _{r, sh, cond} is the specific heat coefficient of the refrigerant; T ₂ is the temperature of the refrigerant flowing into the condenser; T _cond is the superheat temperature of the condenser; T ₃ is the temperature of the refrigerant flowing out of the condenser; r _cond is The structural radius of the plate condenser; A _i,cond is the heat exchange area of the plate condenser; _Tw is the outlet temperature of the hot water side of the condenser; U _sc,cond , U _sh,cond , U _tp,cond are the condensation Comprehensive coefficient of heat transfer in cold water measurement, superheat zone and two-phase zone.

The mathematical model analysis of each component in the above electric heat pump unit has obtained the heat converted by the two circulating media in each component. Each link is related to the complete operation of the electric heat pump through the flow and temperature of the medium, and the parameters that focus on operation control are often the heating capacity, power consumption and energy efficiency ratio of the electric heat pump.

From the formula (2-2), it can be known that the low-level heat energy Q _c collected by the heat collector is mainly determined by the external temperature _To , flow rate and heat collection efficiency of the heat source medium. Assuming that all this heat enters the evaporator, the heat exchange of the heat source medium of the evaporator is equal to the heat energy collected by the heat collector, that is, Q _a =Q _c . The heat Q _eva transferred to the refrigerant in the evaporator is equal to the sum of the heat exchange Q _tp in the two-phase area and the heat exchange Q _sh in the superheated area. , When the heat exchange loss in the two-phase area, the heat exchange of the evaporator satisfies:

Q _eva =Q _tp +Q _sh =Q _a (3)

In addition to the thermal energy input of the low-level heat source, the electric heat pump unit also provides energy for the refrigerant circulation by means of electrical energy. From the previous analysis, the power-consuming components of the electric heat pump system include compressors, fans, and water pumps. Among them, the compressor work has the greatest influence on the heating of the electric heat pump. It can be seen from equation (1) that the electric power P _cmp of the compressor of the electric heat pump will directly affect the flow rate m _r and pressure of the refrigerant in the refrigerant cycle. Equation (2) shows that the flow rate m _r of the refrigerant and the temperature T _cond of the refrigerant are determined by the heat exchange Q _cond of the condenser. Therefore, the final output heat of the electric heat pump should be composed of two parts: the heat energy of the low-level heat source and the electric energy of the electric heat pump compressor. It can be obtained that the heating quantity Q _o output by the electric heat pump, the heat exchange quantity Q _eva of the evaporator and the electric power P _cmp consumed by the compressor have the following equation relationship:

Q _o =Q _eva +P _cmp ·η _cmp (15)

The energy conservation equation describes that the heating capacity Q _o of the electric heat pump contains two parts of energy, the heat absorbed from the outside and the work done by the electric power. In engineering, the concept of energy efficiency ratio is often used when describing the energy-saving performance of temperature control equipment.

COP reflects the energy-saving heating capacity of the electric heat pump. The larger the COP value, the more heat the electric heat pump generates per unit power consumption. From the model analysis of the electric heat pump heating system, it can be known that the COP of the electric heat pump is related to the temperature T _o of the external heat source medium, the heat exchange efficiency of the heat exchanger (evaporator and condenser), and the operating power P of the main power-consuming component (compressor). _cmp related.

According to the model analysis of the units of the electric heat pump heating system, the power-consuming components of the electric heat pump heating system include compressors, fans, and water pumps, among which the compressor is the most power-consuming component in the electric heat pump. The power demand when the electric heat pump starts is mainly from the compressor. In addition, almost all compressors of electric heat pumps use asynchronous motors with good speed regulation characteristics, so the model of asynchronous motors can be used to characterize the power consumption characteristics of electric heat pumps.

Due to the existence of high-order nonlinear flux coupling, the mathematical model of the asynchronous motor in the static coordinate system is relatively complex, and it is often transformed into the d-q two-phase rotating coordinate system. The complex variable model of the d-q coordinate system of the asynchronous motor is described as:

分别定子电压、定子电流、转 子电流、定子磁链和转子磁链的复数向量；R_s、L_s、M、R_r、L_r分别为定子 电阻、定子电感、定转子互感、转子电阻和转子电感；l_s＝L_s-M和l_r＝L_r-M分 别为定子和转子漏电感；ω_r为转子电角速度，单位为rad/s；j是复数单位。In the above formula,

Complex vectors of stator voltage, stator current, rotor current, stator flux linkage and rotor flux linkage, respectively; R _s , L _s , M, R _r , L _r are stator resistance, stator inductance, stator-rotor mutual inductance, rotor resistance and rotor, respectively Inductance; l _s =L _s -M and l _r =L _r -M are the stator and rotor leakage inductances, respectively; ω _r is the rotor electrical angular velocity, in rad/s; j is a complex unit.

From the above formula (4), the matrix equation of the T-type transient equivalent circuit of the induction motor can be obtained:

为感应电动机的速度感应电势。In the formula, p is the differential operator;

Induces an electric potential for the speed of an induction motor.

Because of the existence of _es in the T-type transient equivalent model, the model becomes a complex complex differential equation, which makes the calculation amount of solution and derivation larger and more difficult. Considering that the speed induced electromotive force of the squirrel-cage rotor asynchronous motor is zero ^[40] , that is, _es = 0, it can be further simplified, and an induction motor without speed induced electromotive force is selected for analysis. Its complex variable model can be described as:

T型暂态等值电路都退化为T型稳态等值电路，As shown in Figure 3, in the simplified T-type transient equivalent circuit. The speed induced potential _es in the rotor circuit is ignored, and the rotor equivalent resistance is used instead. When p=jω in steady state operation, the factor

T-type transient equivalent circuits are degenerated into T-type steady-state equivalent circuits.

的乘积构成。然而， 上述电机动态数学模型具有非线性、强耦合的特点，加上微分算子p及启动过程 中转差率s这类不可观测量的存在使得模型求解和参数辨识较为困难且不便于 研究使用。It can be seen that the stator resistance and reactance in the transient T-type equivalent model are consistent with the steady-state model, but the rotor resistance of the transient equivalent model is the steady-state rotor equivalent resistance and differential factor.

The product of . However, the above-mentioned dynamic mathematical model of the motor has the characteristics of nonlinearity and strong coupling, and the existence of unobservables such as differential operator p and slip s in the starting process makes the model solution and parameter identification difficult and inconvenient for research and use.

将上式代入式 (18)可以得到：At the moment when the asynchronous motor starts, because the motor rotor has not yet turned. Therefore, the slip ratio of the asynchronous motor is s=1, and the motor structure can be equivalent to a secondary short-circuit transformer at this time, so there are

Substituting the above formula into formula (18) can get:

In the above formula, L _sσ is the stator leakage inductance coefficient.

将上式代入式(18)可以得到：When the asynchronous machine is running stably with no load, the rotor speed is close to the electromagnetic speed. At this time, the slip s is very small (about 0.03), approximately s=0. At this time, the equivalent model of the motor is similar to the transformer model with an open circuit on the secondary side, there are

Substituting the above formula into formula (18) can get:

The simplified equations (20) and (22) are the port characteristics of the asynchronous motor under the two states of instantaneous startup and no-load stable operation, and the slip s in the transient process from startup to stable operation is from 1 to approximately The change process of 0 is difficult to describe.

Based on the above analysis, the following two aspects of the start-up process of the electric heat pump are studied: the start-up power function fitting based on the test, and the critical start-up voltage analysis based on the model parameter identification.

In order to solve the problem that the slip rate s is difficult to describe and the starting power model cannot be established during the starting process, the test platform makes the electric heat pump start at the rated voltage for many times, and records the current and voltage measurement values. From the measured values of starting current and voltage, some starting power data of the starting process of the electric heat pump can be calculated and drawn into a curve. The test data with higher test accuracy can be selected as the source data for model identification, and based on the source data, the electric heat pump start-up shock multiple (5.9-7.1) and start-up duration (0.22-0.36s) conform to the uniform distribution simulation to enrich the simulation experimental simulation data. Finally, the method of curve fitting is used to identify the starting power model of the electric heat pump.

The quadratic function is used to fit the test data, and after comparing and observing the active power curve, it is found that the curve of the starting power and time of the electric heat pump presents obvious quadratic function characteristics. Therefore, the quadratic function relationship of active power P _s with respect to time t can be constructed as follows:

P _s (t)=a(tb) ² +c (23)

In the above formula, a, b, and c are parameters to be identified. Combining the start-up characteristics of the electric heat pump and the apex characteristic of the quadratic function, the vertex coordinates (t _max , P _max ) of the P _s (t) curve correspond to the time t _max and the peak value P _max when the electric heat pump start-up power reaches the maximum value. Using the experimental data to identify the model parameters, we get:

It is worth noting that: the definition domain of t in the formula should be from the initial start time t ₀ to the time t ₁ when the electric heat pump enters a steady state after the start is completed, and b is approximately half a start cycle. Therefore, the closed-interval functional relationship between the active power P _s and the time t during the starting process of the electric heat pump is:

P _s (t)=-1160(t-0.2) ² +P _max t∈[t ₀ ,t ₁ ] (5)

Table 1 shows the similarity and mean of the quadratic function at each fitting point. The mean similarity of quadratic function to the test power curve fitting of the electric heat pump reaches 0.931.

Table 1

The stator leakage inductance L _sσ , the stator inductance L _s , the stator resistance R _s , and the rotor resistance R _r in the foregoing equations (20) and (22) are unknown parameters to be determined. The parameters can be identified by using the test measurement data of the instant when the asynchronous motor starts and the two periods when the asynchronous motor enters the no-load stable operation.

In order to simplify the complexity of identification, a first-order filter is designed for the differential equations of equations (20) and (22), and the identification of the differential equations is transformed into the least squares identification using the filtered data. And with the help of the test data, the stator and rotor resistance and reactance parameters of the T-type equivalent model are identified, as shown in Table 2.

Table 2

After identifying the impedance parameters of the asynchronous motor, the T-type equivalent circuit can be used to analyze the electromagnetic torque and mechanical torque of the asynchronous motor's starting process. When the rotor of the asynchronous motor rotates, the electromagnetic power PE _absorbed by the rotating rotor from the stator side. Electromagnetic power is applied to the rotor through electromagnetic torque in the magnetic field and makes it rotate, thus completing the conversion process of electrical energy to mechanical energy of the rotor. The electromagnetic torque T _e and P _E have the following relationship:

P _E =ΔP+P _mec =f(U _s ² ,Z) (25)

In the formula, PE is the electromagnetic power of the asynchronous motor; △P is the copper loss of the rotor resistance; P _mec is the total mechanical power generated on the rotating shaft; _{U s} is the terminal voltage of the asynchronous motor; Z is the port equivalent impedance of the asynchronous motor; ω is the per-unit value of the speed of the asynchronous motor.

The electromagnetic torque T _e of the asynchronous motor is related to the speed and terminal voltage, while the mechanical torque T _m of the electric heat pump load is proportional to the square of the speed. The mechanical torque of the electric heat pump load consists of a variable part proportional to the square of the rotational speed and a fixed part determined by friction loss. The mechanical torque T _m of the electric heat pump is:

T _m (ω)=T ₀ +ΔTω ² (27)

In the above formula, T ₀ is the constant coefficient of the fixed part of the mechanical torque; ΔT is the constant coefficient of the proportional change of the mechanical torque about ω ² , take T ₀ =0.8471, ΔT=0.1875 ^[44] .

When the network impedance is not taken into account, the mechanical characteristics of the asynchronous motor (the electromagnetic torque and rotational speed of the asynchronous motor) under different terminal voltages (taking the motor terminal voltages as 1.0pu, 0.98pu, and 0.95pu, respectively) can be obtained based on the values of the above model parameters. The relationship between the curves) is shown in Figure 4.

From the electromagnetic torque and mechanical torque curves of the asynchronous motor, it can be known that when the starting voltage of the motor port is high enough (U ≥ 0.95pu), the electromagnetic torque T _e and the mechanical torque T _m have an intersection point (balanced operating point), at In this case, the asynchronous motor can start and accelerate to a stable operating state; when the starting voltage is lower than a certain threshold (U<0.95pu), there is no intersection between T _e and T _m , and the mechanical torque is always higher than the maximum corresponding to the starting voltage. Electromagnetic torque, the motor cannot be accelerated to a state of stable operation. Therefore, whether there is a balanced operating point between the electromagnetic torque T _e and the mechanical torque T _m is the condition for the successful start of the motor. The peak value of electromagnetic torque is determined by the starting voltage, so whether there is a balanced operating point between T _e and T _m depends on the value of the port voltage when the motor starts.

The voltage value at which there is a unique intersection point between the electromagnetic torque _Te and the mechanical torque T _m of the asynchronous motor is defined as the critical starting voltage U _lim . It can be seen from the formula (26) and Fig. 4 that the maximum electromagnetic torque T _e,max and the maximum rotational speed ω _max of the asynchronous motor are only related to the impedance of the asynchronous motor itself, but have nothing to do with the port voltage. Therefore, as long as the maximum rotational speed ω _max is determined, the critical voltage U _lim at which the electric heat pump load can start normally can be calculated by equation (26) equal to the electromagnetic torque T _e and equation (27) mechanical torque T _m .

T _e (ω _max ,U _lim )=T _m (ω _max ) (28)

In Figure 4, when the port voltage is 0.95pu, there is a unique intersection (T _e,max , ω _max ) between the mechanical torque and the electromagnetic torque of the electric heat pump, that is, its critical starting voltage is 0.95pu. When considering the impedance of the distribution network and the voltage fluctuation of the motor starting, whether the asynchronous motor can run normally depends on whether the voltage value of the node where it is located is lower than the critical voltage U _lim that it can start normally.

2. Identify the state of the electric heat pump based on μPMU measurement

First determine the constraints of the head-end node voltage U _i . According to the existing theory, whether the electric heat pump starts successfully depends mainly on whether the compressor starts successfully. For a fixed-frequency compressor using a single-phase asynchronous motor, the successful start-up depends on whether the port voltage reaches the critical start-up voltage U _lim of the asynchronous motor.

In the actual low-voltage distribution network, the electric heat pumps are scattered all over the power grid, and the voltage of each electric heat pump is different due to the long voltage drop in the distribution line. Therefore, the voltage of the head-end node must be high enough to ensure that the voltage of each electric heat pump after being stepped down by the line is not lower than its critical starting voltage U _lim .

Using the T-type equivalent model of the electric heat pump and the structural characteristics of the parallel connection of the distribution lines with impedance, the equivalent circuit of the network containing n electric heat pumps can be obtained as shown in Figure 5, where U _i is the power supply access point voltage (take the low voltage. Distribution transformer outlet voltage), R _k and X _k (k=1,2,...,n) are the resistance and reactance of the k-th line. R _sk and R _rk are the stator resistance and rotor resistance of the k-th electric heat pump load respectively, X _sk and X _rk are the reactances of the stator and rotor of the k-th electric heat pump load, respectively; X _mk is the excitation reactance; slip _sk = ( ω _s -ω _rk )/ω _s , ω _s is the synchronous angular velocity, and ω _rk is the rotor angular velocity.

In order to obtain the voltage of the kth electric heat pump load access point, the node voltage method can be used to calculate the above-mentioned network equivalent circuit. The figure contains n nodes, which are defined by the node admittance matrix and the characteristics of the equivalent circuit. The node admittance matrix Y of the network can be obtained:

In the above formula, Y _kk is the self-admittance of node k, including the line admittance and the equivalent admittance of the electric heat pump located at node k; Y _{k+1, k} is the mutual admittance between node k+1 and node k, that is is the admittance of line k+1. Since there is only mutual admittance between adjacent nodes, the node admittance matrix Y is a sparse n×n matrix. From this, the node voltage equation Y·U=I in matrix form can be established as follows:

为节点k的节点电压；

为注入节点1的电流且

Z₁为节点1至电源接入点的线路阻抗且Z₁＝R₁+jX₁。在线路阻抗 和电热泵T型等值电路阻抗已知的情况下，根据节点电压方程(30)可以求解 得到节点k的电热泵端口电压

是关于

的线性函数：

In the above formula,

is the node voltage of node k;

is the current injected into node 1 and

Z ₁ is the line impedance from node 1 to the power access point and Z ₁ =R ₁ +jX ₁ . When the line impedance and the T-type equivalent circuit impedance of the electric heat pump are known, the electric heat pump port voltage of node k can be obtained by solving the node voltage equation (30).

its about

The linear function of :

Fig. 6 is the equivalent circuit of the electric heat pump whose terminal voltage is U _k . In order to facilitate the calculation of its electromagnetic torque related to the port voltage and its own impedance, the circuit parts other than the equivalent branch of the rotor impedance are simplified and equivalent. The simplified port equivalent circuit can be obtained as shown in Figure 7.

According to the equivalent principle of Thevenin circuit, the simplified series equivalent potential U _ek and equivalent impedance Re +jX _ek are _calculated to satisfy the following equations:

From the simplified circuit and the new equivalent parameters, the electromagnetic torque T _ek (ω _rk ,U _i ) acting on the motor rotor in the kth electric heat pump load can be obtained as:

In the above formula, _sk and U _ek are the linear functions of the rotor speed ω _rk and the voltage U _i at the head-end node of the platform, respectively. Let the first derivative of equation (33) with respect to the rotor speed ω _rk be 0, then the maximum rotor speed ω _maxrk corresponding to the maximum value of T _ek can be obtained.

It can be seen from equation (34) that the maximum rotor speed ω _maxrk is only affected by the electric heat pump load rotor impedance, stator impedance and excitation impedance and the equivalent impedance of the external network, and has nothing to do with the port voltage. Therefore, after the load impedance and network impedance parameters of each electric heat pump are determined, the maximum rotor speed ω _maxrk of each electric heat pump load can be calculated by formula (34). After obtaining the maximum rotor speed ω _maxrk corresponding to the maximum electromagnetic torque, the mechanical torque of the kth electric heat pump can be obtained according to the relationship between mechanical torque and speed (27) as follows:

In the above formula, T _m (ω) is the mechanical torque of the electric heat pump at the per-unit value ω of the asynchronous motor speed, T ₀ is the constant coefficient of the fixed part of the mechanical torque; △T is the constant coefficient of the mechanical torque proportional to the change of ω ² , usually take T ₀ =0.8471, ΔT = 0.1875. T _k0 and ΔT _k are the constant coefficients of the fixed and variable parts of the mechanical torque of the kth electric heat pump, and their values are the same as those of formula (27). Then, from the equation (28) of the critical startup voltage, combined with the equation (33) and the equation (35), there is T _ek (ω _maxrk , U _i )=T _m (ω _maxrk ). From this equation, the critical value U _lim.k of the voltage U _i at the head-end node of the station area when the electric heat pump located at node k in the low-voltage distribution network can be normally started can be calculated.

To sum up, when there are n electric heat pumps connected to the station area of the low-voltage distribution transformer, the voltage U _i of the head-end node of the station area must satisfy:

U _i ≥{max(U _lim.k )|T _ek (ω _maxrk ,U _lim.k )=T _m (ω _maxrk ),k∈[1,n]} (6)

In the formula, U _i takes the maximum value of the critical value U _lim.k when the electric heat pump load at node k (k=1,2,...,n) can start normally. When U _i satisfies the above formula, even if the voltage of each node in the network drops due to the existence of line impedance and load impedance, it will not be lower than the critical start-up voltage U _lim of the electric heat pump, so it can ensure that each electric heat pump in the network starts successfully.

Micro Phasor Measurement Unit (μPMU) has the advantages of accurate and fast measurement. The present embodiment judges the start-up state of the electric heat pump based on the following two criteria:

Criterion 1. Power Shock

When the electric heat pump is started, there will be a large power shock, because the compressor rotates at zero at the moment of power-on. In order to obtain sufficient electromagnetic torque, the peak value of the starting power of the fixed-frequency electric heat pump is as high as 6 times the rated power, and the duration of the shock process is about 0.3s.

It can also be seen from the quadratic function model of the start-up power of the electric heat pump: the value of the start-up power is large and the growth rate is fast, and the slope of the curve at each moment can be derived from the P _s (t) function (Equation 5), that is, is the rate of change of starting power:

P _s '(t)=-2320(t-0.2) (36)

It can be found that: when t=0.2, the growth rate of the starting power is 0, and the starting power reaches the maximum value. From the quadratic function model of starting power, it can be known that the value range of P _s '(t) in the whole starting process is [-464,464]. Accordingly, it can be identified whether the electric heat pump starting power exists at the current time t according to the power variation and the value of P _s '(t):

In the above formula, P(t) is the load power at the current time t; P(t-1) is the load power at the previous measurement time; N is the measurement frequency at time t, and the power change takes n measurement data The difference is made in turn and the sum is obtained; P'(t) is the rate of change of the load power at the current time t; λ ₁ and λ ₂ are the thresholds corresponding to the criteria. In the presence of μPMU measurement data support, |P′(t)| can be easily approximated:

In the formula, P(t ₂ ) and P(t ₁ ) are the load powers calculated from the voltage and current measurement values of the μPMU device at time t ₂ and t ₁ respectively.

It is worth noting that the values of the thresholds λ ₁ and λ ₂ have a great influence on the identification results. If the values are too small, conventional load fluctuations may be incorrectly identified, and if the values are too large, some impulse power may be missed. Therefore, a reasonable value should not only filter out most of the conventional load fluctuations, prevent misjudgment, but also accurately identify the impact power of the electric heat pump.

Criterion 2. Voltage sag

Based on the start-up test and start-up power model analysis, it is known that the start-up process of the electric heat pump consumes a large amount of power, accompanied by a current inrush, and the large current will cause a significant voltage sag. The voltage sag ΔU has obvious peak and transient characteristics.

The μPMU device can monitor the power flow of the distribution network in real time. When the voltage measurement value U _t at time t is much smaller than the voltage measurement value U _{t-1 at time t-1} , and the voltages at time U _t-1 and t-2 are If the measured value U _t-2 is very close, it can be determined that the voltage sags at time t, so the voltage sag criterion is as follows:

In the formula, the threshold λ ₃ is a large positive number, which is used to describe the degree that U _t is much smaller than U _t-1 ; the threshold λ ₄ is a small positive number, which is used to describe U _t-1 and U _{t- 2} closeness; N is the measurement frequency at time t.

It is worth noting that: when the basic residential load _PL and electric heat pump load with small fluctuation and slow change are included in the network, the startup of a single electric heat pump first affects the rapid rise of the network power, and then causes the voltage to drop significantly: based on the startup test It can be seen that the starting process of the asynchronous motor consumes a lot of power (ΔP=η· _PN , η=5～7), accompanied by current inrush, and the high current causes serious voltage drop. by the voltage sag formula

Due to the large line resistance R of the low-voltage distribution network, the active power P has a great influence on the voltage. The node injection power includes the load steady state power and the electric heat pump startup impulse power, that is, P= _PL +△P. So the voltage sag satisfies:

When k sets of electric heat pumps are started at the same time, then P= _PL + △P·k, resulting in a voltage drop as:

The superposition of the starting power makes the network power climb larger, and at the same time, the voltage drop will be more serious. The value of λ ₁ , λ ₂ , λ ₃ , and λ ₄ is the key to the accuracy of the state criterion identification result. Reasonable selection of the threshold value can accurately identify the impact of the single heat pump startup and the simultaneous startup of multiple heat pumps. Therefore, the threshold values in criteria 1 and 2 can be selected according to the network power and voltage changes when a single heat pump is started, and the selected threshold can more easily judge the impact of multiple heat pumps starting at the same time. However, in the actual power grid, the network power and voltage changes are affected by the network load P= _PL + △P, so the selection of the threshold needs to be analyzed according to the actual measurement data and determined by the method of enumeration.

Next, further identify the number of electric heat pumps that are in the starting state at the current moment. The measured value at a certain time t _n is P(t _n ). If there is a criterion to determine that there is an electric heat pump start-up impulse power at the time t _n , that is to say, the load power contains an instantaneous sudden increase in pulse fluctuation. By comparing the load power before the electric heat pump is started, the electric heat pump startup impulse power P _imp (t _n ) (pulse fluctuation) at the current moment can be separated by the following formula:

P _imp (t _n )=P(t _n )-P(t _m ) (41)

In the formula, P(t _n ) is the measured power value at time t _n ; P(t _m ) is the measured value of power at time t _m when there is no impact power very close to time t _n . When the measurement result does not satisfy the characteristic criterion of the starting state, it is regarded as the moment when there is no impact power. The moment _tm which is closer to the moment _tn is selected to avoid the error caused by the normal fluctuation of the load on the calculation result of Equation (41). Therefore, t _m is taken as the previous moment of the non-impulse power judged by the characteristic judgment of the starting state of the electric heat pump.

The startup of electric heat pumps is often random, so there may be several electric heat pumps in the startup process at the time t _n , then the startup impulse power at the current moment is the superposition of the startup power of several electric heat pumps. From the functional model, it can be concluded that P _imp (t _n ) is the sum of several quadratic functions:

P _imp (t _n )=P _s.1 (t _n )+P _s.2 (t _n )+…+P _si (t _n ) (42)

In the above formula, P _si (t _n ) is the starting power of the ith electric heat pump in the starting state.

Therefore, to identify the number of electric heat pumps that are in the starting state at the current moment is to identify how many quadratic functions are superimposed on P _imp (t _n ). Set the start-up state coefficient K(n)=(k ₁ ,k ₂ ,...,k _N ) as and start-up delay coefficient Δt(n)=(t ₁ ,t ₂ ,...,t _N ) ^T The starting power function P _s (t) of the aforementioned electric heat pump is discretely valued to obtain the starting power function value P _s (t _i ) at different times, and then according to K(n), Δt(n) and P _s ( t _i ) can be combined to describe the superposition of all electric heat pump impulse powers:

S(n)=K(n)·P _s [Δt(n)]=k ₁ ·P _s (t ₁ )+k ₂ ·P _s (t ₂ )+…+k _N ·P _s (t _N ) (43)

In the formula, S(n) is the impulse power of the electric heat pump obtained by the superposition of the quadratic function P _s (t); N is the number of sampling points of the electric heat pump startup power function P _s (t), which can be calculated by combining the data acquisition frequency of the μPMU. Determine; t _N is the starting time corresponding to the starting power P _s (t) sampling point; P _s (t _N ) is the starting power function value at the starting time t _N ; k _N is the number of electric heat pumps at the starting time t _N. When k _N = 0, it means that there is no electric heat pump at the start-up time t _N ; when k _N = k, it means that there are k electric heat pumps at the start-up time t _N , and the start-up power is k·P _s (t _N ) .

To identify the starting number of electric heat pumps, it is necessary to obtain the corresponding starting state coefficient K(n)=(k ₁ , k ₂ when the superimposed power S(n) of the fitting is equal to the impulse power measurement value P _imp (t _n ) ,...,k _N ) ^T . In order to describe the closeness of the measured power P _imp (t _n ) and the superimposed power S(n), the Pearson correlation coefficient γ of P _imp (t _n ) and S(n) can be used as the identification objective function, which defines the equation for:

γ=cov[P _imp (t _n ),S(n)]/[μP _imp (t _n )·μS(n)] (43)

where cov[P _imp (t _n ), S(n)] is the covariance of the electric heat pump impulse power P _imp (t _n ) and the fitted superimposed power S(n), μP _imp (t _n ), μS (n) is the standard deviation corresponding to the measured power and the superimposed power.

The closer the value of γ is to 1, the more similar the two values are, that is, the measured power P _imp (t _n ) and the superimposed power S(n) have a better fit, and the approximate P _imp (t _n )=S( n). At this time, the number of electric heat pumps in the process of starting and the corresponding starting time can be determined by the identified starting state coefficient A(n) and starting delay coefficient Δt(n).

In the above identification process using the Pearson correlation coefficient γ as the objective function, it is necessary to optimize the open state coefficient K(n)=(k ₁ , k ₂ ,...,k _N ), because the measurement data of μPMU Having a higher density makes where the value of N is larger, while the element k _N is a discrete non-negative integer. Therefore, the above function problem is a discrete combinatorial optimization problem, which can be solved quickly with the help of mathematical optimization software (such as CPLEX).

3. Orderly electricity consumption method of electric heat pump based on measurement and identification

From the point of view of direct load control, it is often necessary to convert the voltage problem into a load power problem. Therefore, all nodes connected to the electric heat pump load in the station area are equivalent to a generalized node m, as shown in Figure 8. The radial low-voltage distribution network can be regarded as a two-node network after being equivalent, the first node is connected to the power grid, and the second node is connected to the load in the station area.

Analyze the current relationship between the head-end node n of the station area and the load generalized node m, and its value satisfies:

In the formula, δ is the phase angle difference between the voltage of the m node and the voltage of the n node. The complex power of node n can be expressed as:

为电流相量

的共轭；

为阻抗角。由于低压配电网的线路较短、 线路阻抗R较大且相角差δ较小，所以由式(44)、(45)可以化简有功功率P和 电压U_n之间的关系如下：In the formula,

is the current phasor

the conjugation of ;

is the impedance angle. Since the line of the low-voltage distribution network is short, the line impedance R is large, and the phase angle difference δ is small, the relationship between the active power P and the voltage _Un can be simplified by equations (44) and (45) as follows:

It can be seen that the injected power P of the head-end node n and the node voltage U _n approximately satisfy a linear variation relationship. Therefore, when the voltage exceeds the limit, active power can be used to adjust the voltage. The linear variation coefficient k _p can be easily calculated from the measurement data of the μPMU:

In the formula, _Un (t ₀ ) and P(t ₀ ) are the current node voltage and power measured at time t ₀ ; P(t ₁ ) and _Un (t ₁ ) are the current node power measured at time t ₁ and voltage. The difference from the identification criterion is that there should be a longer time interval between time t ₀ and time t ₁ . The ratio of the difference between the measured values can be approximated as the slope k _p of the linear relationship, so the voltage and injected active power of the head-end node n are obtained as follows:

U _n =k _P ·[PP(t ₀ )]+U _n (t ₀ ) (9)

In the formula, U _n is the voltage required by the head-end node n; P is the injected power required by the head-end node n. When one of the quantities is known, the value of the other can be calculated. For the node connected to the electric heat pump, in order to meet the starting voltage requirement of the electric heat pump, the value of the node voltage _Un is linear from the critical starting voltage, that is, the process of solving P from the determined _Un .

In order to ensure the voltage constraint U _lim of the head-end node of the station area, it is required that the injected power does not exceed P _lim =f(U _lim ). Therefore, the shortage of the operating power and power limit of the electric heat pump is the amount of power reduction required to meet the start-up of the electric heat pump, and then the problem of the starting voltage of the electric heat pump is transformed into the problem of the load power constraint in the platform area.

Combining the above steps, considering the impact power of electric heat pump startup, the strategy flow of using μPMU measurement to identify the startup state of electric heat pump and ensure the critical voltage of each node by limiting the load of the station area is shown in Figure 9. In practical applications, there are two different situations where the load power exceeds the limit in the station area: (1) the superimposed peak value of the start-up impulse power of the electric heat pump started at the same time exceeds the limit; (2) the load of the steady-state operation of the distribution transformer is large. The power limit is exceeded due to the superposition of partial starting power. The methods for orderly use of electric heat pumps in response to these two different situations are as follows:

After the required power reduction amount P _cat is obtained from the measurement criterion and the limit of the threshold voltage, the corresponding control strategy can be formulated to respond. The operating characteristics of the electric heat pump mainly depend on the compressor, and the power consumption during the start-up process is several times that of the normal operation. Therefore, different from the steady-state operation strategy of the electric heat pump, the following startup strategy will be the whole process of the electric heat pump operation. It is divided into three states: startup, running, and shutdown. Therefore, the relationship between the power characteristics of the electric heat pump and the room temperature during the operation of the electric heat pump when the impulse power is considered is shown in Figure 10.

In Figure 10, T _max and T _min are the upper and lower temperature limits of the indoor temperature range, respectively. As long as the strategy of controlling the electric heat pump can ensure that the room temperature where it is located does not exceed the temperature limit, the user's comfort requirement can be satisfied. . Under the condition that the electric heat pump load accounts for a large proportion, there are a large number of users who randomly turn on and use the electric heat pump, and the operation of each electric heat pump changes with the external environment and user needs. If the temperature change curve in Figure 6 is subdivided into several operating areas section, then at any time, each electric heat pump may be in any section of the control cycle, and the power consumed includes both steady-state operation power and start-up impact power. Therefore, the orderly power consumption strategy of the electric heat pump should consider not only the steady-state operating power but also the start-up impulse power. That is to control the electric heat pump in the starting state and the running state at the same time.

Based on the above considerations, the electric heat pumps are grouped into load groups of corresponding states according to their working states. The overall operating power of the load state group can be changed by formulating a control strategy to make the electric heat pumps in different working states convert each other. At any time in the distribution network, there are three different states of electric heat pumps: startup, operation and shutdown. The electric heat pumps are grouped using the following load state groups:

In the formula, t is the current moment of strategy execution; S _t , R _t and B _t are the load groups of the electric heat pump currently working on startup, operation and shutdown, and the corresponding loads are n ₁ , n ₂ and n ₃ , the electric heat pump load The total number n=n ₁ +n ₂ +n ₃ . It can be seen that with the working of the electric heat pump load, the working state of the electric heat pump load and the numbers n ₁ , n ₂ and n ₃ in S _t , R _t and B _t will change. Therefore, the set of electric heat pumps in all states in the system can be expressed as:

Different electric heat pump loads in different working states consume different electric power at the current time t. Similarly, the power set of equipment group is defined according to the method of load group set:

为电热泵在时刻t所消耗的电功率，不同工作状态下的电热泵对应 于不同负荷功率(启动功率P_s(t)、额定功率P_N和零)；D_t内设备在t时刻从电网 中汲取的功率可表示为：In the formula,

is the electric power consumed by the electric heat pump at time t, and the electric heat pump in different working states corresponds to different load powers (starting power P _s (t), rated power P _N and zero); the equipment in D _t is removed from the power grid at time t The power drawn can be expressed as:

为在t时刻处于启动状态的第k台电热泵启动功率；

为在t时刻 处于运行状态的第j台电热泵额定功率。因此有序用电控制策略对响应群体中的 电热泵进行温度设定值调整或开关操作，进而改变电热泵负荷运行功率，通过响 应群体中各个电热泵的功率改变，可以影响电热泵负荷群的整体运行功率。In the formula,

is the starting power of the kth electric heat pump in the starting state at time t;

is the rated power of the jth electric heat pump in operation at time t. Therefore, the orderly power consumption control strategy adjusts the temperature setting value or switches the electric heat pump in the response group, and then changes the load operating power of the electric heat pump. By changing the power of each electric heat pump in the response group, it can affect the electric heat pump load group. overall operating power.

After obtaining the operating power of the electric heat pump and the shortfall of the load power limit P _lim as the power reduction amount P _cat in the strategy flow chart 9, the key to formulating the control strategy is to choose how many electric heat pumps in which working state are converted into other electric heat pumps. A working state to change the overall operating power of the load state group. After the working state of the electric heat pump changes, the process that will cause the power to change includes: from running to off state, from off to on state. The parameter changes corresponding to the state change are shown in Table 3.

table 3

In the formulation of the orderly power consumption strategy, according to the above two state conversion processes, the electric heat pumps in different working states are transformed into each other to change the overall operating power of the load state group to respond to the power regulation of the system. The implementation steps of the orderly electricity consumption strategy are as follows:

1) According to the measurement and identification methods, the load information is obtained and the load groups in different states are determined.

2) Prioritize the electric heat pump loads of the two equipment groups in startup and operation states.

In view of the above-mentioned two different situations of load power exceeding the limit, different methods of orderly use of electric heat pumps are specifically adopted:

Scenario 1: Controlling the startup sequence, limiting startup power

When multiple electric heat pumps are started at the same time, the superimposed start-up power impact may exceed the load power limit P _lim of the station area. When the load of the distribution network is not heavy, the start-up impact power of some heat pumps will not cause the load of the station area to exceed the limit. Therefore, artificially controlling the start-up sequence of the electric heat pumps is a way to avoid their simultaneous start-up and avoid serious power shocks. . In order to determine the starting sequence of the electric heat pump, the priority is arranged according to the critical starting voltage of the electric heat pump and the voltage level of the access node, and the number of starting units is determined according to the priority and search conditions.

First, according to the relationship between the average voltage U _k of the node where the electric heat pump is located in the historical running time from small to large, the electric heat pumps that are in startup and about to be started are sorted. The principle that the electric heat pump with the critical voltage is started first) makes the electric heat pump with a larger connected voltage U _k have a later order. Then each electric heat pump generates a steady-state voltage priority F _u according to the deviation of U _k from the rated voltage U _N and its critical starting voltage U _lim as follows. The larger the value of F _u , the higher the priority, the same below.

Then, according to the indoor temperature T _k where the electric heat pump N _k is located, the electric heat pumps with different grades of room temperature are generated according to the principle of "low temperature, start first" (that is, the electric heat pumps with room temperature close to the lower limit of the temperature range are preferentially started) to generate different temperature priorities F. _T is as follows:

Combining the above sorting and priority calculation methods, the electric heat pump start-up sequence that has been sorted according to the voltage average value U _k in the historical running time is updated. The update method is to add the corresponding priority to the current sequence position number. If the electric heat pump N _k is ranked in the k position according to the average voltage U _k in the historical operation time, the starting sequence number after considering the priority is:

K′=k+F _u + _FT (56)

Scenario 2: Controlled orderly shutdown, reducing steady-state power

When the load of the distribution transformer in steady-state operation is large, the superimposed part of the starting power may cause the load power to exceed the limit in the station area. Similar to controlling the start-up sequence of the electric heat pump, in order to reasonably reduce the steady-state operation load and reduce the total load of the station area, it is necessary to reasonably arrange the number and sequence of shutdown units. Priority is arranged according to the critical starting voltage of the electric heat pump and its indoor temperature level, and the number of units to be turned off is determined according to the priority and search conditions.

First, according to the relationship between the indoor temperature T _in where the electric heat pump is located from large to small, the electric heat pumps in stable operation are sorted according to the principle of "high temperature, turn off first" (that is, the electric heat pump whose room temperature is close to the upper limit of the temperature range is preferentially turned off). , so that the electric heat pump whose room temperature is close to the lower limit of the temperature range has a later order. Then, according to the indoor temperature T _j where the electric heat pump N _j is located, different temperature priorities F′ _T are generated for the electric heat pumps with different grades of room temperature as follows:

Then according to the principle of "easy to start, first to close" (that is, the electric heat pump with the applied voltage higher than the starting critical voltage is preferentially turned off), according to the difference between the average applied voltage U _j and its critical starting voltage U _lim of the electric heat pump N _j in the long-term operation state. The deviation generation voltage priority F' _U is as follows:

The update method of the shutdown sequence of the electric heat pump is similar to the update method of the startup sequence of the electric heat pump. The current sequence position number is added with the corresponding priority. If the electric heat pump N _j is located in the j position according to the U _j order, the shutdown sequence number after considering the priority is : j'= _j +F'T+ _F'U .

3) Determine the order and quantity of controlled electric heat pump equipment. When the system wants to limit the load starting power, the load group S _t is controlled to limit the number of starting equipment; when the system wants to reduce the load to ensure the start-up is completed, the load group R _t is controlled to reduce the steady-state load. The number of electric heat pumps to be controlled is determined by the power reduction amount P _cut required by the system, and the method for determining the number of response equipment is shown in Table 4.

Table 4

For two different cases of the startup power shock problem, a conditional search is performed on the startup load group S _t and the operating load group R _t based on this table rule. From the above search direction and search conditions, the groups R' _t and S' _t that finally participate in load control at time t can be obtained as follows.

Turn off the equipment in R' _t or turn on the equipment in S' _t in sequence according to the needs, in which delay intervention is performed on the start-up time of the electric heat pump unit in the start-up sequence S' _t to ensure that the start-up time is staggered as much as possible to avoid shock superposition. The delay time of the heat pump to be started by the ith unit is Δt=i·t' (i=1,2,...,k ₁ ), and t' is the maximum start-up time of the electric heat pump. The operating state of these devices can be directly changed through the start-stop control of the electric heat pump. The purpose of limiting the number of starting units to avoid the superposition of starting power exceeding the limit and reducing the steady-state load to preferentially meet the starting demand.

Claims (10)

1. a large-scale electric heat pump orderly control method based on μPMU device, is characterized in that, comprises the following steps: 1) Modeling of electric heat pump heating system; 2) Identify the state of the electric heat pump based on μPMU measurement; 3) Orderly electricity consumption strategy of electric heat pump based on measurement and identification. 2. The method for controlling orderly electricity consumption of a large-scale electric heat pump based on a μPMU device according to claim 1, wherein in the step 1) in the electric heat pump heating system The functional relationship between the electric power P _cmp of the compressor and the rotational speed N and the refrigerant flow rate m _r :

In the above formula, c ₁ is the suction specific volume of the compressor, n is the variable index of the compression process, η _cmp is the ratio of the indicated power of the compressor to the power consumption,

Represents the ratio of exhaust pressure to suction pressure; The heat transfer model of the condenser is as follows:

In the formula, C _{r, sh, cond} is the specific heat coefficient of the refrigerant; T ₂ is the temperature of the refrigerant flowing into the condenser; T _cond is the superheat temperature of the condenser; T ₃ is the temperature of the refrigerant flowing out of the condenser; r _cond is The structural radius of the plate condenser; A _i,cond is the heat exchange area of the plate condenser; _Tw is the outlet temperature of the hot water side of the condenser; U _sc,cond , U _sh,cond , U _tp,cond are the condensation is the comprehensive coefficient of heat transfer in cold water measurement, superheated zone and two-phase zone of the boiler, and _Tw is the temperature of hot water; The heat exchange of the evaporator satisfies: Q _eva =Q _tp +Q _sh =Q _a In the above formula, Q _eva is the heat transferred to the refrigerant in the evaporator, and Q _a is the heat absorbed by the evaporator from the heat source medium. The complex variable model of the d-q coordinate system of the asynchronous motor is described as:

In the above formula,

Complex vectors of stator voltage, stator current, rotor current, stator flux linkage and rotor flux linkage, respectively; R _s , L _s , M, R _r , L _r are stator resistance, stator inductance, stator-rotor mutual inductance, rotor resistance and rotor, respectively Inductance; l _s =L _s -M and l _r =L _r -M are the stator and rotor leakage inductances, respectively; ω _r is the rotor electrical angular velocity, in rad/s; j is a complex unit. 3. the orderly control method of large-scale electric heat pump based on μPMU device according to claim 1, is characterized in that, in described step 1), the electric heat pump starting process active power P _s of electric heat pump heating system is about time The closed interval functional relationship of t is: P _s (t)=-1160(t-0.2) ² +P _max t∈[t ₀ ,t ₁ ] In the above formula, t ₀ is the initial start time, t ₁ is the time when the electric heat pump enters a steady state after starting up, and P _max is the maximum starting power of the electric heat pump. 4. the method for orderly control of large-scale electric heat pump based on μPMU device according to claim 1, it is characterized in that, in described step 2), first establish head-end node voltage U _i constraint, the station head-end node voltage U _i must meet: U _i ≥{max(U _lim.k )|T _ek (ω _maxrk ,U _lim.k )=T _m (ω _maxrk ),k∈[1,n]} In the above formula, U _lim.k is the critical value when the electric heat pump load at node k can start normally, T _ek (ω _maxrk , U _lim.k ) is the electromagnetic torque acting on the motor rotor in the kth electric heat pump load, ω _maxrk is the maximum rotor speed when T _ek takes the maximum value, U _lim.k is the critical starting voltage at node k, and T _m (ω _maxrk ) is the mechanical torque of the electric heat pump at the per-unit value ω _maxrk of the asynchronous motor speed. 5. the orderly control method of large-scale electric heat pump based on μPMU device according to claim 4, is characterized in that, according to power shock and voltage sag to judge electric heat pump start-up state, and wherein power shock criterion is as follows:

In the above formula, P(t) is the load power at the current time t; P(t-1) is the load power at the previous measurement time; N is the measurement frequency at time t, and the power change takes n measurement data in turn Differences and summations are obtained; P'(t) is the rate of change of load power at the current time t; λ ₁ and λ ₂ are the thresholds corresponding to the criteria. The voltage sag criterion is:

In the formula, λ ₃ and λ ₄ are the thresholds used to describe the proximity of U _t-1 and U _t-2 , and N is the measurement frequency at time t. 6. The orderly control method for large-scale electric heat pump based on μPMU device according to claim 1, characterized in that, in the step 2), the number of quadratic functions in the electric heat pump startup impulse power is utilized to identify The number of electric heat pumps in the activated state. 7. The orderly control method of large-scale electric heat pump based on μPMU device according to claim 1, is characterized in that, in described step 3), load voltage is converted into load power, and is specifically as follows: The voltage and injected active power of the head-end node n satisfy: U _n =k _P ·[PP(t ₀ )]+U _n (t ₀ ) In the formula, U _n is the voltage required by the head-end node n; P is the injected power required by the head-end node _n , Un (t ₀ ), P (t ₀ ) are the current node voltage and power measured at time t ₀ k _p is the slope, and when one of the quantities is known, the value of the other quantity can be calculated. 8. The method for controlling orderly power consumption of a large-scale electric heat pump based on a μPMU device according to claim 7, wherein the orderly power consumption strategy implementation steps are as follows: 1) Obtain load information according to measurement and identification methods and determine load groups in different states; 2) Prioritize the electric heat pump loads of the two equipment groups in startup and running states; 3) Determine the order and quantity of controlled electric heat pump equipment. 9. The method for controlling the orderly power consumption of a large-scale electric heat pump based on a μPMU device according to claim 8, wherein in the step 2), the superimposed peak value of the start-up impulse power of the electric heat pump started at the same moment exceeds the limit , add the steady-state voltage priority, different temperature priority and the current sequence position number to get the startup sequence number. 10. The method for orderly power consumption control of a large-scale electric heat pump based on a μPMU device according to claim 8, wherein in the step 2), when the load for the steady-state operation of the distribution transformer is large, the superimposed part of the starting power The resulting power exceeds the limit, the current sequence position number, the priority of the electric heat pump sorted by the average value of the applied voltage, and the priority of different temperatures are added to obtain the shutdown sequence number.

Images