Abstract : Through the analysis of the energy consumption of the traction system of Guangzhou Metro Line 1 trains during normal operation, it is shown that the energy-saving effect of regenerative braking of metro trains is obvious. Under the condition of sufficiently high operating density, the energy consumed by the braking resistor is very limited. Keywords : Metro train; energy consumption; regenerative braking; braking resistor; energy saving 0 Introduction In recent years, with the rapid development of China's economic level, the metro system in major cities is developing rapidly. In the next few years, more metro lines and more metro trains will be put into operation in China. While convenient urban rail transit has brought great convenience to citizens' travel, it has also led to a rapid increase in electricity consumption. As we all know, the rapid development of modern economy must rely on energy, and China is a country with relatively scarce energy. Therefore, studying the energy consumption of metro trains and analyzing and researching energy-saving methods for metro trains is an urgent task. 1 Use of Regenerative Braking in Metro Trains The characteristics of urban rail transit trains are short station spacing and frequent starting and braking during train operation. In the case of Guangzhou Metro, the existing lines basically brake very quickly when the train reaches its maximum speed. Meanwhile, to ensure trains operate accurately according to the timetable, urban rail vehicles use cruise mode in ATO (Automatic Train Operation) mode. Currently, most metro trains in my country use DC power supply via overhead contact lines/rails, while traction systems are mostly AC drive systems with variable voltage and frequency conversion. Trains absorb energy from the power grid during traction and use regenerative braking to feed the braking energy back to the grid during braking. When the grid voltage rises to a certain level (1800V), resistance braking is employed. Given the characteristics of metro vehicles—rapid starting, rapid braking, and precise preset speeds across the entire line—trains consume a large amount of electrical energy during starting and inevitably generate considerable braking energy during braking. Regenerative braking converts kinetic energy into electrical energy, which is then fed into the grid for use by other trains, significantly reducing the actual energy loss of the train. However, due to the influence of the train timetable and various factors along the entire line, braking resistors are configured on the trains to absorb braking energy that other trains on the line cannot absorb during braking. The usage of braking resistors on the four existing metro lines in Guangzhou is shown in Table 1. The use of braking resistors has the following drawbacks: 1) Braking resistors consume electrical energy, wasting braking energy; 2) For braking resistors with forced air cooling, the train must provide forced air, which is also a waste of electrical energy; 3) Onboard braking resistors increase the weight of the train and its cost; 4) The heat generated by the braking resistors affects other equipment on the train and in the tunnel. Based on these drawbacks, some have proposed using supercapacitors to replace braking resistors in recent years. Supercapacitors are also considered in two ways: onboard and ground-based. Whether it is necessary to use supercapacitors in subway trains depends on factors such as energy efficiency and manufacturing and maintenance costs. To obtain accurate data on train energy consumption, we measured the energy consumption of Guangzhou Metro Line 1 trains during normal operation. Analysis of the measurement data clarified the actual energy consumption of Guangzhou Metro Line 1 trains, providing a basis for comprehensively considering energy-saving solutions for subway trains. 2. Data Acquisition 2.1 Data Acquisition Time (See Table 2) Since the train intervals vary at different times of operation, and the absorption of feedback electrical energy by other trains during braking may differ at different intervals, we measured the energy consumption of the trains during peak and off-peak operation. The Guangzhou Metro Line 1 train traction control unit (DCU) signal processing board has corresponding signal measurement interfaces, which greatly facilitated our measurement. In Table 3, A327 refers to a signal processing board of the DCU, and PB115 refers to the measurement interface. Since the signals we collected are ultimately used for train control vector calculation, the signal values ​​obtained here are very accurate. Therefore, measurement errors are not considered. [b]3. Concepts Used in Calculation[/b] Given that we have not previously seen a complete measurement, calculation, and analysis of metro train energy consumption, this paper defines the following concepts, mainly to clearly describe the energy consumption of the trains. 1) Train traction system input energy: The total electrical energy obtained by the traction system from the power grid when the train is running in the section. Its value is the integral of the product of capacitor voltage and forward line current over the running time. 2) Regenerative braking power fed into the grid: The energy fed back into the power grid by the train through regenerative braking when running in the section. Its value is the integral of the product of capacitor voltage and reverse line current over the braking time. 3) Braking resistor energy consumption: The energy consumed in the braking resistor when the braking chopper phase is activated. Its value is the integral of the product of capacitor voltage and chopper phase current over the braking time. 4) Actual train traction power: The electrical energy actually consumed by the traction system when the train is running in the section. Actual train traction power = Train traction system input energy - Regenerative braking power fed into the grid. In short, during train traction, the train absorbs energy from the power grid; this is the "train traction system input energy." When the train brakes, provided the electric braking meets the braking requirements, the train feeds electrical energy back to the power grid; this is the "regenerative braking power grid input energy." During train braking, if the grid voltage rises above 1800V due to the train's feedback braking, the train can no longer continue feedback braking. At this point, the train's braking resistor is activated, and the electrical energy consumed by the braking resistor is the "braking resistor energy consumption." During train operation on the main line, regardless of whether the electrical energy is consumed during acceleration or during braking and is consumed by the braking resistor, it is still considered energy consumption for the power grid. Therefore, we define the concept of "actual train traction electrical energy." From the perspective of the power grid's energy output to the train, "actual train traction electrical energy" includes "braking resistor energy consumption." To gain a more intuitive understanding of where the electrical energy goes when the train is running on a section of track, we use the following two concepts for analysis. 1) The ratio of regenerative braking energy fed into the grid to the total input energy of the train traction system is expressed by the following formula: When a subway train accelerates, a large amount of electrical energy is converted into kinetic energy by the traction system. During braking, a significant portion of this kinetic energy is also fed back to the grid as electrical energy through the traction inverter. By quantitatively comparing and analyzing the two quantities, "regenerative braking energy fed into the grid" and "train traction system input energy," we can intuitively understand the energy transfer and consumption process during traction, where the traction system absorbs grid energy, converts it into kinetic energy, and then converts it back into electrical energy during braking. For the grid, "train traction system input energy" is not the true consumption; the actual traction energy of the train is the true consumption of grid energy. In other words, "train traction system input energy" minus the "regenerative braking energy fed into the grid" represents the true consumption of grid energy by the train traction system, i.e., "actual traction energy of the train." 2) The ratio of braking resistor energy consumption to actual train traction power is expressed by the following formula: Ratio of braking resistor energy consumption to actual train traction power = During the acceleration phase, the train traction system absorbs a large amount of power from the grid, and during the regenerative braking phase, it feeds some power back to the grid. In addition, from the grid's perspective, all the power that cannot be fed back is absorbed by the train; this portion is what is referred to as "actual train traction power." However, for the train itself, most of this power is consumed by the entire system during various operating phases such as acceleration, coasting, and braking. This includes track resistance, wind resistance, heat consumption of the entire traction system (traction inverter, traction motor, track), and energy losses during energy conversion. These energy consumptions are related to the track the train travels on, the overall network's operating status, and the efficiency of the train traction system; we will not analyze them in detail here. A small portion of the "actual traction power" is generated when the grid voltage rises above 1800 V, at which point the train's braking resistors are put into operation. The "braking resistor energy consumption" at this time helps us quantitatively analyze and discuss the value of the train's braking resistors. Therefore, we quantitatively discuss the "braking resistor energy consumption" by using the "ratio of braking resistor energy consumption to actual traction power." [b]4 Energy Consumption Test Data[/b] Guangzhou Metro Line 1 consists of 6-car trains, 4 motor cars and 2 trailer cars. Considering that the traction systems of all 4 motor cars in a train are identical, this measurement only collected signals from the traction system of one motor car. Through calculation of the measurement data, we obtained the energy consumption of the train during peak hours in each section, summarized in Tables 4 and 5. The section numbers in the tables are sequentially ordered starting from the originating station. During peak hours, when the train is traveling uphill, the ratio of regenerative braking power fed into the grid to the total input energy of the train's traction system is 0.524, and the ratio of braking resistor energy consumption to the actual traction energy is 0.0834. When the train is traveling downhill, the ratio of regenerative braking power fed into the grid to the total input energy of the train's traction system is 0.496, and the ratio of braking resistor energy consumption to the actual traction energy is 0.0008. Tables 6 and 7 summarize the energy consumption of trains operating on different sections during off-peak hours. During off-peak hours, when the train is traveling uphill, the ratio of regenerative braking power fed into the grid to the total input energy of the train's traction system is 0.47, and the ratio of braking resistor energy consumption to the actual traction energy is 0.0009. When the train is traveling downhill, the ratio of regenerative braking power fed into the grid to the total input energy of the train's traction system is 0.42, and the ratio of braking resistor energy consumption to the actual traction energy is 0.0312. For safety reasons, Guangzhou Metro Line 1 activates a chopper to discharge the capacitor when the train stops and doors open. Analysis of the signals collected from the train shows that all records of braking resistor energy consumption less than 0.01 kW/h occurred at the moment the train stopped and doors opened. Therefore, records less than 0.01 kW/h represent energy discharged from the capacitor, not energy consumed during braking. [b]5 Measurement Results Analysis[/b] 1) The actual number of times the resistor brake operates is very low, regardless of whether the train is running during peak or off-peak hours. During peak hours, there is only one significant resistor braking event in the 13th section of the upward direction, consuming 10.5 kW/h of energy. During off-peak hours, there is only one significant resistor braking event in the 5th section of the downward direction, consuming 3.1 kW/h of energy. Figure 2 shows the traction braking signal of the train in the 6th section of the upward direction during off-peak hours; the three curves from top to bottom are XIN, XUD, and XIBS. Since there is no current in the chopper phase during the entire traction and braking process, the train does not undergo resistance braking during operation in this section. Simultaneously, as shown in the curve in the figure, the capacitor voltage drops rapidly while the chopper phase current increases instantaneously; this is the record of the chopper phase activating at the moment the doors open. 2) Currently, approximately 48% of the "train traction system input energy" on Guangzhou Metro Line 1 is fed back to the grid for consumption by other trains during braking, and approximately 2.9% of the "actual traction energy" is consumed by the braking resistor. 97.1% of the "actual traction energy" is consumed during train operation due to track resistance, wind resistance, and the supplementary air braking during braking. Therefore, regardless of the energy-saving method used, the theoretically possible energy savings will not exceed 2.9% of the current "actual traction energy." 3) The highest ratio of "braking resistor energy consumption to actual train traction power" occurs during peak-hour travel (upward), reaching 8.34%. During peak-hour downward travel and off-peak-hour upward travel, the braking chopper phase is not activated. During off-peak downward travel, the ratio of "braking resistor energy consumption to actual train traction power" is 3.12%. The total ratio of "braking resistor energy consumption to actual train traction power" for both round trips is 2.9%. Figure 3 shows the traction braking signals for the 13th section of the upward travel during peak hours. It can be seen that the current in the braking resistor is very large when the chopper phase is activated, but this situation rarely occurs. The feedback power during train braking is approximately 48% of the actual traction power. Although the timing of starting and braking of trains operating on the main line is affected by multiple factors, we can simplify it. Let the total number of trains operating on the main line be m, and the number of braking trains be n. Then the equation for the braking energy of the braking trains being exactly absorbed by the traction trains is as follows: n × 0.48 = (m - n) × 1. The result is: n / m = 0.67 That is, as long as the number of braking trains on the entire line does not exceed 67% of the number of operating trains, and without considering line losses, the energy feedback from the braking trains can be completely absorbed by the trains that are currently accelerating. In this case, the regenerative braking of the trains will not cause an increase in grid voltage, and the train braking resistors will not be used. In actual operation, it is rare for more than 67% of trains to brake simultaneously, so the need for resistor braking is also relatively rare. Of course, this analysis does not consider trains coasting or stopped; it is merely a simplified analysis. 4) Because passenger volume is significantly higher during peak hours than during off-peak hours, passenger volume affects traction energy consumption. During off-peak hours, train intervals increase and passenger flow decreases, but the use of braking resistors does not increase. That is, as long as the train schedule is well organized, the train density is high enough, the intervals are uniform, and the timing of starting, accelerating, and braking of trains within the same power supply section is properly coordinated, regenerative braking can always achieve good results. 6 Conclusion Currently, the energy-saving effect of regenerative braking in subway trains is significant, and the energy consumed by the braking resistors is limited. This is why some Japanese urban rail systems do not have braking resistors. Of course, the subway system is a very complex project. The limited electrical energy consumed by the braking resistor does not mean that the braking resistor is not very useful, nor does it mean that the braking resistor can be eliminated. Whether to use a braking resistor and whether to install the braking resistor on the train must be considered comprehensively in the design of the subway system. This article analyzes the electrical energy consumption of subway trains and clarifies the actual situation of the energy consumption of Guangzhou Metro Line 1 trains, which is of reference value for comprehensively considering the absorption device of braking electrical energy of subway trains. References : [1] Siemens. Description of Guangzhou Metro Line 1 vehicles [M]. Berlin: Siemens, 1998.