Negative sequence inverse time limit protection curve correction method based on thermal network method in negative sequence current abrupt change scene

Abstract

A negative sequence inverse time limit protection curve correction method based on a thermal network method in a negative sequence current abrupt change scene comprises the following steps: taking the heating process of a generator into consideration, and performing node division on each part of the generator; calculating the thermal resistance of the convection heat dissipation coefficient; constructing a thermal network equation based on steady-state loss, and solving the steady-state operation temperature of each node; calculating the heat capacity of each node, constructing a heat network differential equation set, and solving the transient operation temperature by using a Runge-Kutta method; and inverse time limit protection curve correction adapting to negative sequence current abrupt change is carried out. According to the method, the adaptability of protection to current abrupt change is improved. And the operation benefit of the unit is improved to the greatest extent while the safe operation of the unit is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of negative-sequence inverse time protection, and in particular to a negative-sequence inverse time protection curve correction method based on a thermal network method in a negative-sequence current mutation scenario. Background Art

[0002] Constrained by the reality of reverse distribution of energy and load in my country, long-distance high-voltage transmission systems are used to connect provincial power grids. After a single-phase grounding fault occurs, due to factors such as large single-machine capacity and low system equivalent impedance, the negative sequence current intruding into the generator is large, and a current component with double frequency is induced on the rotor surface, causing the rotor surface temperature to rise rapidly in a short period of time, and the generator is cut off through the negative sequence inverse time protection. The high-voltage transmission system is equipped with a comprehensive reclosing device, which trips the single phase after the fault and enters the non-full-phase operation stage. After a period of time, the protection single phase is reclosed, and the power supply can be quickly restored in the case of a transient fault. Under the influence of negative sequence current, the negative sequence protection operates before the single-phase reclosing, resulting in the failure of the reclosing. Therefore, it is necessary to suppress the negative sequence current intruding into the generator and extend the negative sequence inverse time protection action time. This requires the ability to accurately characterize the negative sequence inverse time protection action time, determine whether the negative sequence suppression strategy is effective, and make timely adjustments to the negative sequence suppression strategy. However, the conventional negative-sequence inverse time protection used in the generator regards the rotor heating as an adiabatic process, which is obviously not in line with the actual situation. After the negative-sequence current is suppressed and decreased, the heat accumulation rate decreases, the influence of the heat dissipation system is significantly enhanced, the actual protection action time is delayed, and the error caused by the conventional negative-sequence inverse time protection is large.

[0003] At present, there are two main methods for improving conventional negative sequence inverse time protection. One is to fit the negative sequence limit bearing capacity curve of the generator to achieve the matching of the protection curve with the negative sequence limit bearing capacity curve. However, this method is not universal and it is difficult to achieve the precise matching of the two curves. The other is to study the heat accumulation process of the generator and quantify the role of influencing factors such as the heat dissipation system to correct the negative sequence inverse time protection curve. However, both methods failed to break the static description relationship between the negative sequence current and the action time, and therefore could not accurately reflect the temperature change of the rotor. Summary of the invention

[0004] In order to solve the problem that conventional negative-sequence inverse time protection of generators is difficult to accurately characterize the protection action time in the negative-sequence current mutation scenario. The present invention provides a negative-sequence inverse time protection curve correction method based on the thermal network method in the negative-sequence current mutation scenario. By studying the heating process of the generator, comprehensively considering the heating conditions of different nodes of the generator and the influence of factors such as rotor rotation on the convective heat transfer coefficient, the thermal energy model of the generator is constructed using the thermal network method, and the accurate calculation of the rotor surface temperature is achieved. This method improves the adaptability of the protection to current mutations. While ensuring the safe operation of the unit, the operating efficiency of the unit is maximized.

[0005] The technical solution adopted by the present invention is:

[0006] A method for correcting a negative-sequence inverse time protection curve based on a thermal network method in a negative-sequence current mutation scenario comprises the following steps:

[0007] Step 1: Consider the heating process of the generator and divide the nodes of each component of the generator;

[0008] Step 2: Calculate the thermal resistance of the convection heat dissipation coefficient;

[0009] Step 3: Construct a thermal network equation based on steady-state loss and solve the steady-state operating temperature of each node;

[0010] Step 4: Calculate the heat capacity of each node, construct the thermal network differential equations, and use the Runge-Kutta method to solve the transient operating temperature;

[0011] Step 5: Modify the inverse time protection curve to adapt to the sudden change of negative sequence current.

[0012] In step 1, the heating process of the generator is considered, the part that is greatly affected by the negative sequence current is divided into a node separately, and the negative sequence inverse time protection curve is corrected based on the temperature rise of the node, and the remaining nodes are divided according to the various components of the generator; specifically as follows:

[0013] In the context of large-scale high-voltage interconnected transmission systems, after a single-phase grounding fault occurs, the negative sequence current intruding into the generator is large and induces a double-frequency current component in the rotor. Affected by the skin effect, the double-frequency component is concentrated on the rotor surface and flows in the path formed by the wedge, tooth, and fixed ring, such as Figure 1 As shown in Figure 2, affected by the contact resistance and high current density, the equivalent heating resistance value corresponding to the double frequency component is relatively large, and the slot wedge can be heated to the softening point in a short time. Therefore, the wedge node with the most serious heat generation in the rotor is divided separately, and the time required for the rotor to reach the limit operating temperature is measured based on the wedge node temperature. The node division of the entire generator is shown in Figure 2.

[0014] Since the number of nodes directly affects the rate of subsequent temperature solutions, the more nodes there are, the slower the solution rate. At the same time, the accuracy requirements of temperature solutions are taken into account. The stator and rotor surfaces with the most serious heat generation are divided into wedge nodes (composed of wedges and rotor tooth surfaces), and the remaining parts of the generator correspond to one node each, such as the casing, end cover, etc. In addition, since the air gap space and the end space gas in the generator do not directly interact with the external environment, and the convection heat dissipation coefficient changes under the influence of rotor rotation and forced heat dissipation, it is divided into two independent nodes for heat exchange. Finally, the external environment is divided into a reference node to determine the temperature changes of the remaining nodes.

[0015] In step 2, the convection heat transfer coefficient between the air gap 4 and the end space node is corrected, the conduction thermal resistance or convection heat transfer thermal resistance between each node is calculated, and the steady-state loss of the generator is calculated; specifically as follows:

[0016] In the actual operation of the generator, the influence of radiation heat transfer on heat transfer is extremely small and can be ignored. Therefore, when constructing the thermal energy model of the generator, only the conduction thermal resistance or convection heat transfer thermal resistance between each node is calculated. The calculation formula is as follows:

[0017]

[0018] Among them, R c is the conduction thermal resistance; R d is the convection heat transfer thermal resistance; l is the distance between the two nodes; A c is the heat conduction area; A d is the heat dissipation area; c is the thermal conductivity of the material; d is the convective heat transfer coefficient.

[0019] The convective heat transfer coefficient is affected by factors such as surface shape, fluid properties and fluid state. The air gap and end space fluid of the generator are affected by factors such as rotor rotation, and the fluid state changes. Therefore, the convective heat transfer coefficient needs to be recalculated. The calculation formula is as follows:

[0020]

[0021] Among them, λ gap is the air gap convection heat transfer coefficient; end is the end space convection heat transfer coefficient; Nu is the Nusselt number; k air is the thermal conductivity of air; D h is the hydraulic diameter of the air gap; v end is the cooling medium flow rate in the end space.

[0022] The steady-state heating of the generator is mainly caused by copper loss, iron loss and mechanical loss. The corresponding calculation formula is as follows:

[0023]

[0024] Among them, P Cu is the copper loss; m is the number of phases of the stator and rotor windings; I is the effective value of the current flowing through the windings; R is the equivalent resistance of the stator and rotor windings; P Fe is the iron loss; k e is the correction factor; p k is the iron loss per unit mass; p h is the hysteresis loss per unit mass; p e is the eddy current loss per unit mass; m Feis the core mass in the calculation area; f is the frequency of the alternating magnetic field; B is the magnetic flux density; σ h is the hysteresis loss coefficient; σ e is the eddy current loss coefficient; P w is the iron loss; k is the surface roughness coefficient; C f is the friction coefficient; ρ is the air density; ω is the rotor angular velocity; r is the rotor radius; l is the rotor length; P b is the iron loss; C b is the bearing coefficient; D m is the bearing diameter.

[0025] In step 3, the thermal network equation is constructed using the thermoelectric analogy principle to solve the steady-state operating temperature of each node; specifically, as follows:

[0026] When the generator is in normal operation, affected by copper loss, iron loss, and mechanical loss, its temperature rises slowly, and under the influence of heat dissipation factors, it finally reaches a stable operating state. By solving the steady-state operating temperature of each node at this time, it can be used as the initial value for calculating the node temperature after the fault. Since the thermal network method has the property of thermoelectric analogy, the node voltage method for solving the electrical network is introduced for solution. The calculation formula is as follows:

[0027]

[0028] Where Y represents the branch thermal admittance matrix; represents the node power matrix; represents the node temperature matrix; A represents the branch correlation matrix; A T Represents the transposed matrix of the branch incidence matrix.

[0029] The branch correlation matrix is ​​formed by stipulating the positive direction of heat transfer between nodes, as follows:

[0030] For nodes with heat conduction, the heat conduction direction is artificially specified to form a directed graph with n nodes and b branches. The branch association matrix A is defined as [a ik ], 1≤i≤n, 1≤k≤b, where row numbers correspond to nodes and column numbers correspond to branches. If branch k is associated with node i and its direction leaves node i, then a ik =1; if branch k is associated with node i and its direction points to node i, then a ik = -1; if branch k is not associated with node i, then a ik =0;

[0031] Calculate the thermal resistance between each node to form the branch thermal admittance matrix, as follows:

[0032] In step 2, the thermal resistance between the related nodes is obtained, that is, the thermal resistance of each branch R1, R2, ..., R b, then the branch thermal admittance matrix can be expressed as

[0033] Substituting the power vector composed of the power of each node into equation (14), the steady-state operating temperature of each node can be solved as follows:

[0034] The steady-state loss power P1, P2, ..., P of the corresponding components of each node is calculated by the steady-state loss formula of the generator. n , forming a node power matrix The branch association matrix A, branch thermal admittance matrix Y and node power matrix Substituting into equation (16), the node temperature matrix can be solved Among them, θ1, θ2, …, θ n They are the steady-state operating temperatures of the components corresponding to nodes 1 to n respectively.

[0035] In step 4, after a single-phase grounding fault occurs, the steady-state operating temperature of each node is taken as the initial value, the power of each node is corrected by measuring the negative sequence current parameter, and the temperature of each node is estimated; specifically, as follows:

[0036] After the negative sequence current invades the generator, the rotor temperature rises rapidly. It is necessary to monitor the transient operating temperature of each node to prevent the rotor from being damaged by heat accumulation. The thermal capacity is introduced to construct a thermal network differential equation group. The steady-state operating temperature of each node is used as the initial value. The power of each node is corrected by measuring the negative sequence current parameters to achieve the solution of the transient operating temperature.

[0037] The thermal network differential equations of the i-th node can be expressed as:

[0038]

[0039] Among them, C i represents the heat capacity of the i-th node; θ i represents the transient temperature of the i-th node; P i represents the corrected power of the i-th node; G ii represents the self-thermal conductance of the ith node, which is the sum of all thermal conductivities connected to the ith node; G ij represents the mutual thermal conductivity of the i-th node, whose value is the thermal conductivity between the i-th node and the j-th node; n represents the number of nodes.

[0040] When the number of nodes is large, the solution speed is slow and cannot meet the real-time requirements of temperature monitoring. Therefore, the Runge-Kutta method is introduced to quickly solve the thermal network differential equation of the i-th node as formula (12), and the transient operating temperature is calculated as shown in formula (13):

[0041]

[0042] In formula (12), f i () represents the relationship between the differential of the transient temperature of the i-th node with respect to time and the transient temperature of each node; θ1,θ2,…,θ i ,…,θ n-1 Represent the transient temperatures of node 1 to node n-1 respectively.

[0043]

[0044] In formula (13), m represents the number of iterations; n represents the number of nodes; h represents the calculation step size; θ i(m+1) represents the transient temperature of node i obtained at the m+1th iteration; θ i(m) represents the transient temperature obtained at the mth iteration of node i; k i1 , k i2 , k i3 , k i4 They represent the change slope coefficients required for computing node i at the m+1th iteration; θ 1(m) ,θ 2(m) ,…,θ n-1(m) They represent the transient temperatures obtained from node 1 to node n-1 at the mth iteration; k 11 , k 21 ,…,k (n-1)1 They represent the slope coefficients of the transient temperature update from node 1 to node n-1; k 12 , k 22 ,…,k (n-1)2 They represent the slope coefficients of the transient temperature update from node 1 to node n-1; k 13 , k 23 ,…,k (n-1)3 They respectively represent the change slope coefficients obtained by updating the transient temperature from node 1 to node n-1.

[0045] In step 5, after the negative sequence current is suppressed and mutated, the temperature of each node before the mutation is taken as the initial value, and the power of each node is corrected respectively with a set of preset negative sequence current values, and the temperature is calculated. When the wedge node temperature reaches the limit temperature, the protection action point corresponding to each negative sequence current preset value is recorded, and the negative sequence inverse time protection curve is refitted; the details are as follows:

[0046] Since the conventional negative sequence inverse time protection curve regards the rotor heating as an adiabatic process for parameter setting, it ignores the heat transfer from the rotor to the surrounding components and the influence of the cooling system, which is obviously not in line with the actual situation. In order to extend the action time of the negative sequence inverse time protection during non-full-phase operation and avoid the risk of group shedding caused by the protection before the reclosing action, it is necessary to suppress the negative sequence current invading the generator during this period. The resulting negative sequence current mutation further affects the rotor heat accumulation process, resulting in deviations in the description of the action time of the conventional negative sequence inverse time protection, which needs to be corrected, including the following steps:

[0047] S5.1: After the negative sequence suppression strategy is put into use, the temperature estimated at each node before the negative sequence current mutation is recorded;

[0048] S5.2: Correct the power loss of each node by using the negative-sequence power corresponding to each preset negative-sequence current prediction value;

[0049] S5.3: Taking the temperature of each node before the negative sequence current mutation as the initial value, the corrected power vector is brought into the thermal network differential equation group to calculate the temperature of each node. When the wedge node temperature reaches the limit temperature, the protection action point corresponding to each negative sequence current effective value is recorded;

[0050] S5.4: Fit each protection action point to obtain the modified negative-sequence inverse time protection action curve.

[0051] The present invention provides a negative sequence inverse time protection curve correction method based on a thermal network method in a negative sequence current mutation scenario, and the technical effects are as follows:

[0052] 1) This invention considers the heating process of the generator, constructs a generator thermal energy model based on the thermal network method, and realizes the solution of the steady-state operating temperature and transient operating temperature of each node; based on the generator thermal energy model, it realizes the accurate characterization of the rotor surface temperature, and then proposes a correction scheme for the negative sequence inverse time protection curve under the negative sequence current mutation scenario. The adaptability of the negative sequence inverse time protection is improved, and the operating efficiency of the unit is maximized while ensuring the safe operation of the unit.

[0053] 2) The present invention analyzes the influence of negative sequence current on the heating of the rotor surface, finely divides the nodes of the components with serious heating, and realizes the accurate calculation of the convective heat transfer thermal resistance by correcting the convective heat transfer coefficient of the air gap and the end space. Therefore, a more accurate generator thermal energy model is constructed, and the local heating of the rotor can be monitored.

[0054] 3) The present invention introduces temperature into the negative-sequence inverse time protection through the generator thermal energy model, which breaks through the precise description of negative-sequence current and action time of conventional negative-sequence inverse time protection and enhances the adaptability of negative-sequence inverse time protection.

[0055] 4) Based on the thermal energy model of the generator, the present invention proposes a correction scheme for the negative-sequence inverse time protection curve. The corrected protection curve is consistent with the actual negative-sequence limit bearing capacity of the generator, which greatly improves the unit efficiency while ensuring the safe and stable operation of the generator. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] The present invention will be further described below in conjunction with the accompanying drawings and embodiments:

[0057] Figure 1 Flow chart of the method of the present invention.

[0058] Figure 2(a) shows the generator structure and the negative sequence current flow. Figure 1 ;

[0059] Figure 2(b) is a second schematic diagram of the generator structure and negative sequence current flow;

[0060] FIG. 2( c ) is an enlarged view of point A in FIG. 2( a ).

[0061] Among them: 1-stator, 2-rotor, 3-shaft, 4-air gap, 5-stator yoke, 6-rotor winding, 7-rotor teeth, 8-wedge, 9-fixed ring, 10-cooling medium, I0-negative sequence current.

[0062] Figure 3 Schematic diagram of generator node partitioning.

[0063] Figure 4 Schematic diagram of the generator thermal energy simulation model.

[0064] Figure 5 is the estimated error of the node steady-state operating temperature.

[0065] Figure 6(a) shows the transient temperature estimation error of the wedge node (rotor).

[0066] Figure 6(b) shows the transient temperature estimation error of the wedge node (stator).

[0067] Figure 7 This is a comparison diagram between the conventional negative-sequence inverse time protection curve and the protection curve that is redrawn after the negative-sequence current changes suddenly.

[0068] Figure 8 To redraw the negative-sequence inverse time protection curve after the negative-sequence current changes suddenly. DETAILED DESCRIPTION

[0069] A method for correcting the negative-sequence inverse-time protection curve based on the thermal network method in the scenario of negative-sequence current mutation. Based on the thermal network method, the temperature rise of the negative-sequence current invading the generator is considered, and the nodes of the generator components are divided based on the degree of heat generation; the influence of factors such as rotor rotation is taken into account, the convection heat transfer coefficient is corrected, and then the thermal resistance between each node is calculated, thereby forming a complete generator thermal energy model; based on the thermal energy model, a thermal balance equation is constructed to achieve the solution of the steady-state and transient operating temperature of each node, and the Runge-Kutta method is introduced to improve the solution speed of the transient operating temperature to meet the protection speed requirements; further proposed is an inverse-time protection curve correction scheme for negative-sequence current mutation based on the real-time temperature of the wedge node. The adaptability of the negative-sequence inverse-time protection to current mutations and the unit operation efficiency are improved. Specifically, the following steps are included:

[0070] Step 1: After the negative sequence current invades the generator, a counter-rotating magnetic field is generated in the air gap, and a harmonic component is induced on the rotor surface. Under the influence of the skin effect, the harmonic component migrates to the teeth and wedges on the rotor surface and flows in the axial path composed of teeth, wedges, and fixed rings, as shown in Figure 2. The high-density harmonic component generated on the rotor surface significantly increases its corresponding equivalent resistance, and there is a large amount of contact resistance in its flow path. Therefore, under the influence of the negative sequence current, the rotor surface temperature will rise rapidly, and local high temperature may cause damage to the rotor surface. Therefore, the rotor surface temperature needs to be monitored in real time.

[0071] Since the thermal network method can subdivide the heat-generating object into multiple nodes for temperature estimation according to actual needs, it can greatly improve the calculation speed while ensuring the calculation accuracy, so it can be used for thermal analysis of generators. Since the surface temperature of the rotor is the most heated, it is necessary to divide it into separate nodes for temperature estimation, and use the node temperature as the basis to measure whether the generator has reached the limit state. In order to reduce the complexity of subsequent calculations, the remaining nodes are divided according to the generator components, thereby obtaining the topological structure of the thermal network, such as Figure 3 shown.

[0072] Step 2: In the thermal network method, the heat transfer process between nodes is characterized by thermal resistance, so the thermal resistance between nodes needs to be calculated. Since radiation heat transfer has little effect on the heating process of the generator, only conduction heat resistance and convection heat transfer heat resistance are considered when calculating thermal resistance.

[0073] Heat conduction in the generator can be divided into two types: flat plate heat conduction and circular tube heat conduction. The thermal resistance of the flat plate conduction can be expressed as:

[0074]

[0075] Where L is the distance between two nodes, i.e. the material thickness; λ c A is the thermal conductivity of the material;c is the heat conduction area.

[0076] The circular tube thermal resistance can be expressed as:

[0077]

[0078] Among them, r1 is the inner radius of the tube; r2 is the outer radius of the tube; λ e A is the thermal conductivity of the material; e is the heat conduction area.

[0079] The thermal resistance of convection heat transfer can be expressed as:

[0080]

[0081] Among them, λ d is the convective heat transfer coefficient; A d It is the contact area between the component and the cooling medium.

[0082] Convective heat transfer coefficient λ d Affected greatly by the surface shape, fluid properties and flow state, during the actual operation of the generator, the rotation of the rotor will cause Taylor vortices to be generated inside the air gap. This vortex flow directly affects the convective heat transfer coefficient of the heat dissipation medium, thereby affecting the heat transfer effect inside the motor. Similar situations also occur in the end space of the generator, so it is necessary to correct the convective heat transfer coefficient between the air gap and the end space.

[0083] The flow state of air in the air gap can be judged by Taylor number:

[0084]

[0085] Where δ is the air gap width; D is the rotor outer diameter; v gap is the gas flow velocity inside the air gap; γ is the kinematic viscosity of air.

[0086] The Taylor number is then used to calculate the Nusselt number under different flow conditions:

[0087]

[0088] Where Pr is the Prandtl number.

[0089] Then the convective heat transfer coefficient in the air gap can be expressed as:

[0090]

[0091] Among them, k air is the thermal conductivity of air; D h is the hydraulic diameter of the air gap.

[0092] Since the air flow state in the end space is relatively complex, in order to reduce the calculation complexity, the empirical formula combining natural convection and forced convection can be used to calculate the convection heat transfer coefficient:

[0093] h end =6.22v end +41.4(7);

[0094] Among them, v end is the gas velocity in the end space.

[0095] Step 3: After obtaining the topological structure and thermal resistance parameters of the thermal network, the thermal network equation can be constructed to solve the steady-state operating temperature of each node, as shown in equation (8).

[0096]

[0097] Where Y represents the branch thermal admittance matrix; R1, R2, ..., R b are the thermal resistance of each branch respectively; represents the node power matrix; P1, P2, ..., P n are the power of each node respectively; represents the node temperature matrix; θ1, θ2, …, θ n are the steady-state operating temperatures of each node; A represents the branch association matrix; b is the number of branches; and n is the number of nodes.

[0098] When the generator is in stable operation, the heat generated by the generator is mainly caused by copper loss, iron loss and mechanical loss. Therefore, the power of each node is also calculated according to the above losses. The corresponding calculation formula is as follows:

[0099]

[0100] Among them, P Cu is the copper loss; m is the number of phases of the stator and rotor windings; I is the effective value of the current flowing through the windings; R is the equivalent resistance of the stator and rotor windings; P Fe is the iron loss; k e is the correction factor; p k is the iron loss per unit mass; p h is the hysteresis loss per unit mass; p e is the eddy current loss per unit mass; m Fe is the core mass in the calculation area; f is the frequency of the alternating magnetic field; B is the magnetic flux density; σ h is the hysteresis loss coefficient, σ e is the eddy current loss coefficient; P w is the iron loss; k is the surface roughness coefficient; C f is the friction coefficient; ρ is the air density; ω is the rotor angular velocity; r is the rotor radius; l is the rotor length; P b is the iron loss; Cb is the bearing coefficient; D m is the bearing diameter.

[0101] Calculate the corresponding steady-state power loss of each node and substitute it into formula (8) to calculate the steady-state operating temperature of each node, which is used as the initial value for calculating the transient operating temperature of each node after a fault.

[0102] Step 4: After the fault occurs, the steady-state operating temperature of each node of the generator is taken as the initial value. By measuring the negative sequence current parameters and correcting the power of each node, the temperature of each node is calculated in real time by constructing a group of thermal network differential equations. The thermal network differential equation of the i-th node can be expressed as:

[0103]

[0104] Among them, C i represents the heat capacity of the i-th node; θ i ′ represents the transient temperature of the i-th node; P i ′ represents the corrected power of the i-th node; G ii represents the self-thermal conductance of the ith node, which is the sum of all thermal conductivities connected to the ith node; G ij represents the mutual thermal conductivity of the i-th node, whose value is the thermal conductivity between the i-th node and the j-th node; n represents the number of nodes.

[0105] The heat capacity of each node can be expressed as:

[0106] C i =ρ i c i V i (11);

[0107] Among them, ρ i , c i , V i They represent the density, specific heat capacity and volume corresponding to the i-th node respectively.

[0108] The transient temperature of each node after the fault can be obtained by solving the thermal network differential equation. However, due to the large number of nodes, that is, the large number of unknowns in the differential equation group, the solution time is long and cannot meet the real-time requirements of temperature estimation. Therefore, the Runge-Kutta numerical solution is introduced for rapid solution. The thermal network differential equation of the i-th node is rewritten as formula (12), and the transient temperature is estimated as shown in formula (13):

[0109]

[0110] Among them, m represents the number of iterations; n represents the number of nodes; and h represents the calculation step size.

[0111] Step 5: After the negative sequence current is suppressed and decreased, the heat accumulation rate slows down, and the influence of the heat dissipation system is significantly enhanced. Conventional negative sequence inverse time protection regards rotor heating as an adiabatic process, which leads to deviations in the protection action time characterization. Therefore, it needs to be corrected. The steps are as follows:

[0112] S5.1: After the negative sequence suppression strategy is put into use, the temperature estimated at each node before the negative sequence current mutation is recorded;

[0113] S5.2: Correct the power loss of each node by using the negative-sequence power corresponding to each preset negative-sequence current prediction value;

[0114] S5.3: Taking the temperature of each node before the negative sequence current mutation as the initial value, the corrected power vector is brought into the thermal network differential equation group to calculate the temperature of each node. When the wedge node temperature reaches the limit temperature, the protection action point corresponding to each negative sequence current effective value is recorded;

[0115] S5.4: Fit each protection action point to obtain the modified negative-sequence inverse time protection action curve.

[0116] Example:

[0117] First, build a generator thermal energy model in the power system simulation software. Its structure is as follows: Figure 4 As shown in the figure, the ambient temperature is set to 300K and the rotor temperature rise limit is set to 420K.

[0118] (1) Node temperature prediction based on thermal network method:

[0119] Solving the thermal network equation and the thermal network differential equation can obtain the steady-state temperature and transient temperature of each node, avoiding the investment of additional temperature measurement equipment. Among them, the solution of transient temperature depends on the steady-state operating temperature of each node, so it is necessary to verify whether the solution error of the steady-state operating temperature of each node is within the allowable range, that is, whether it affects the subsequent solution of transient operating temperature.

[0120] Set a sufficiently long simulation time and compare the steady-state operation of each node obtained by the generator thermal energy simulation model with the results of solving the thermal network equations, such as Figure 5 The calculation error of the steady-state operating temperature of each node is shown in Table 1.

[0121] Table 1 Steady-state operating temperature estimation error of each node

[0122]

[0123] It can be seen that the steady-state operating temperature of each node obtained by solving the thermal network equation is very close to the actual value, and the error is within the controllable range, so it can meet the accuracy requirements of the subsequent transient temperature solution.

[0124] After a single-phase grounding fault occurs, negative sequence current invades the generator, causing serious heating of the rotor surface temperature. Therefore, it is necessary to focus on monitoring the temperature of the wedge node to prevent the rotor surface temperature from being too high and causing rotor damage. The effective value of the negative sequence current invading the generator after a single-phase grounding fault is set to 500A, and the single-phase protection tripping takes 0.06s. Affected by the single-phase tripping, the negative sequence current will reach 450A during non-full-phase operation. The comparison between the predicted temperature and the actual temperature of the stator wedge node and the rotor wedge node is shown in Figure 6 (a) and Figure 6 (b), and the error of the remaining nodes is shown in Table 2.

[0125] Table 2 Estimated error of transient operating temperature at each node

[0126]

[0127]

[0128] It can be seen that solving the transient temperature of each node through the thermal network differential equation can adapt to the sudden change of negative sequence current and has high prediction accuracy, which is not affected by the prediction time.

[0129] (2) Characterization of negative sequence inverse time protection curve based on thermal network method:

[0130] After the negative sequence current suddenly changes, the conventional negative sequence inverse time protection will act in advance because it regards the rotor heating as an adiabatic state, making it difficult to accurately describe the impact of the negative sequence current. The conventional negative sequence inverse time protection curve is compared with the protection curve re-characterized after the negative sequence current suddenly changes. Figure 7 shown.

[0131] The root mean square error between the negative-sequence inverse time protection curve characterized by the thermal network method and the negative-sequence limit bearing capacity curve is 0.1152K, and the two curves are highly overlapped, indicating that the negative-sequence protection curve characterization method proposed in the present invention can meet the accurate calculation of the rotor temperature, and can maximize the unit efficiency while ensuring the correct protection action.

[0132] However, the general negative sequence inverse time protection curve is limited by the number of setting parameters and cannot match the negative sequence limit bearing capacity curve. In the negative sequence current range of [150,500], the general negative sequence inverse time protection acts in advance and cannot give full play to the unit benefits; when the negative sequence current is less than 150A, the general negative sequence inverse time protection curve and the negative sequence limit bearing capacity curve show a cross trend, which also means that when the negative sequence current is small, the general negative sequence inverse time protection will delay the action, bringing great safety hazards to the unit.

[0133] Set different negative sequence suppression levels to redraw the negative sequence inverse time protection curve after the negative sequence current suddenly changes, such as Figure 8 shown.

[0134] It can be seen that the negative sequence current mutation has a greater impact on the protection. As the negative sequence current suppression ratio gradually increases, the heat accumulation rate slows down, the influence of the heat dissipation system is significantly enhanced, the actual protection action is prolonged, and the protection curve shows an upward trend. If the negative sequence inverse time protection is not corrected, the protection will be triggered in advance, the unit benefits cannot be fully utilized, and the safe and stable operation of the power system will be affected. It can be seen that the negative sequence inverse time protection curve correction scheme proposed in the present invention can adapt to the negative sequence current mutation situation.

Claims

1. A negative sequence inverse time protection curve correction method based on thermal network method under negative sequence current mutation scenario, characterized in that The following steps are involved: Step 1: Consider the heating process of the generator and divide the nodes of each component of the generator; Step 2: Calculate the thermal resistance of the convection heat dissipation coefficient; Step 3: Construct a thermal network equation based on steady-state loss and solve the steady-state operating temperature of each node; Step 4: Calculate the heat capacity of each node, construct the thermal network differential equations, and use the Runge-Kutta method to solve the transient operating temperature; Step 5: Modify the inverse time protection curve to adapt to the sudden change of negative sequence current.

2. According to claim 1, a negative sequence inverse time protection curve correction method based on a thermal network method in a negative sequence current mutation scenario is characterized by: In step 1, the heating process of the generator is considered, the part greatly affected by the negative sequence current is divided into a node separately, and the negative sequence inverse time protection curve is corrected based on the temperature rise of the node, and the remaining nodes are divided according to the components of the generator.

3. According to claim 2, the method for correcting a negative sequence inverse time protection curve based on a thermal network method in a negative sequence current mutation scenario is characterized by: The wedge nodes in the rotor where heat is most severe are divided separately, and the time required for the rotor to reach the limit operating temperature is measured based on the temperature of the wedge nodes.

4. According to claim 1, the method for correcting a negative sequence inverse time protection curve based on a thermal network method in a negative sequence current mutation scenario is characterized by: In step 2, the convection heat transfer coefficient between the air gap 4 and the end space node is corrected, the conduction thermal resistance or convection heat transfer thermal resistance between each node is calculated, and the steady-state loss of the generator is calculated; specifically as follows: When constructing the generator thermal energy model, the conduction thermal resistance or convection heat transfer thermal resistance between each node is calculated. The calculation formula is as follows: Among them, R c is the conduction thermal resistance; R d is the convection heat transfer thermal resistance; l is the distance between the two nodes; A c is the heat conduction area; A d is the heat dissipation area; c is the thermal conductivity of the material; d is the convective heat transfer coefficient; The convective heat transfer coefficient is affected by factors such as surface shape, fluid properties and fluid state. The air gap and end space fluid of the generator are affected by the rotor rotation factor, and the fluid state changes. Therefore, the convective heat transfer coefficient is recalculated. The calculation formula is as follows: Among them, λ gap is the air gap convection heat transfer coefficient; end is the end space convection heat transfer coefficient; Nu is the Nusselt number; k air is the thermal conductivity of air; D h is the hydraulic diameter of the air gap; v end is the cooling medium flow rate in the end space; The steady-state heating of the generator is caused by copper loss, iron loss and mechanical loss. The corresponding calculation formula is as follows: Among them, P Cu is the copper loss; m is the number of phases of the stator and rotor windings; I is the effective value of the current flowing through the windings; R is the equivalent resistance of the stator and rotor windings; P Fe is the iron loss; k e is the correction factor; p k is the iron loss per unit mass; p h is the hysteresis loss per unit mass; p e is the eddy current loss per unit mass; m Fe is the core mass in the calculation area; f is the frequency of the alternating magnetic field; B is the magnetic flux density; σ h is the hysteresis loss coefficient; σ e is the eddy current loss coefficient; P w is the iron loss; k is the surface roughness coefficient; C f is the friction coefficient; ρ is the air density; ω is the rotor angular velocity; r is the rotor radius; l is the rotor length; P b is the iron loss; C b is the bearing coefficient; D m is the bearing diameter.

5. According to claim 1, the method for correcting a negative sequence inverse time protection curve based on a thermal network method in a negative sequence current mutation scenario is characterized by: In step 3, the thermal network equation is constructed using the thermoelectric analogy principle to solve the steady-state operating temperature of each node; specifically, as follows: When the generator is in normal operation, affected by copper loss, iron loss, and mechanical loss, its temperature rises slowly, and under the influence of heat dissipation factors, it finally reaches a stable operating state. The steady-state operating temperature of each node at this time is used as the initial value for calculating the node temperature after the fault. Since the thermal network method has the property of thermoelectric analogy, the node voltage method for solving the electrical network is introduced for solution. The calculation formula is as follows: Where Y represents the branch thermal admittance matrix; represents the node power matrix; represents the node temperature matrix; A represents the branch correlation matrix; A T represents the transposed matrix of the branch incidence matrix; The branch correlation matrix is ​​formed by stipulating the positive direction of heat transfer between nodes, as follows: For nodes with heat conduction, the heat conduction direction is specified to form a directed graph with n nodes and b branches. The branch association matrix A is defined as [a ik ], 1≤i≤n, 1≤k≤b; Calculate the thermal resistance between each node to form the branch thermal admittance matrix, as follows: In step 2, the thermal resistance between the related nodes is obtained, that is, the thermal resistance of each branch R1, R2, ..., R b , then the branch thermal admittance matrix is ​​expressed as The steady-state loss power P1, P2, ..., P of the corresponding components of each node is calculated by the steady-state loss formula of the generator. n , forming a node power matrix The branch association matrix A, branch thermal admittance matrix Y and node power matrix Substituting into equation (16), the node temperature matrix can be solved Among them, θ1, θ2, …, θ n They are the steady-state operating temperatures of the components corresponding to nodes 1 to n respectively.

6. According to claim 1, a negative sequence inverse time protection curve correction method based on a thermal network method in a negative sequence current mutation scenario is characterized by: In step 4, after a single-phase grounding fault occurs, the steady-state operating temperature of each node is taken as the initial value, the power of each node is corrected by measuring the negative sequence current parameter, and the temperature of each node is estimated; specifically as follows: The thermal capacity is introduced to construct the thermal network differential equations. The steady-state operating temperature of each node is taken as the initial value. The power of each node is corrected by measuring the negative sequence current parameter to solve the transient operating temperature. The thermal network differential equations of the i-th node are expressed as: Among them, C i represents the heat capacity of the i-th node; θ i represents the transient temperature of the i-th node; P i represents the corrected power of the i-th node; G ii represents the self-thermal conductance of the ith node, which is the sum of all thermal conductivities connected to the ith node; G ij represents the mutual thermal conductivity of the i-th node, and its value is the thermal conductivity between the i-th node and the j-th node; n represents the number of nodes; The Runge-Kutta method is introduced to quickly solve the thermal network differential equation of the i-th node and rewrite it into equation (12). Then the transient operating temperature is calculated as shown in equation (13): In formula (12), f i () represents the relationship between the differential of the transient temperature of the i-th node with respect to time and the transient temperature of each node; θ1,θ2,…,θ i ,…,θ n-1 Respectively represent the transient temperatures of nodes 1 to n-1; In formula (13), m represents the number of iterations; n represents the number of nodes; h represents the calculation step size; θ i(m+1) represents the transient temperature of node i obtained at the m+1th iteration; θ i(m) represents the transient temperature obtained at the mth iteration of node i; k i1 , k i2 , k i3 , k i4 They represent the change slope coefficients required for computing node i at the m+1th iteration; θ 1(m) ,θ 2(m) ,…,θ n-1(m) They represent the transient temperatures obtained from node 1 to node n-1 at the mth iteration; k 11 , k 21 ,…,k (n-1)1 They represent the slope coefficients of the transient temperature update from node 1 to node n-1; k 12 , k 22 ,…,k (n-1)2 They represent the slope coefficients of the transient temperature update from node 1 to node n-1; k 13 , k 23 ,…,k (n-1)3 They respectively represent the change slope coefficients obtained by updating the transient temperature from node 1 to node n-1.

7. According to claim 1, a negative sequence inverse time protection curve correction method based on a thermal network method in a negative sequence current mutation scenario is characterized by: The step 5 comprises the following steps: S5.1: After the negative sequence suppression strategy is put into use, the temperature estimated at each node before the negative sequence current mutation is recorded; S5.2: Correct the power loss of each node by using the negative-sequence power corresponding to each preset negative-sequence current prediction value; S5.3: Taking the temperature of each node before the negative sequence current mutation as the initial value, the corrected power vector is introduced into the thermal network differential equation group to calculate the temperature of each node. When the wedge node temperature reaches the limit temperature, the protection action point corresponding to each negative sequence current effective value is recorded; S5.4: Fit each protection action point to obtain the modified negative-sequence inverse time protection action curve.

Images