Power supply system and method of multipath alternate pre-charge

Abstract

The application relates to the field of body robot power supply, and provides a multi-path alternate pre-charging power supply system and a multi-path alternate pre-charging method. The method comprises the following steps: system initialization, controlling all pre-charging selection switches and all main power supply switches to be in an off state; in a preset sequence, a first pre-charging selection switch is turned on, so that a common pre-charging resistor charges a load side capacitor of the path in current limiting, and at most only one pre-charging selection switch is in a conductive state at the same time; the voltage of the load side of the path is detected in real time through the voltage detection circuit of the path, when the voltage of the load side of the path reaches a preset threshold, the main power supply switch of the path is closed, and the pre-charging selection switch of the path is disconnected, so as to release the common pre-charging resistor; in the preset sequence, a next pre-charging selection switch is turned on, and the detection of the voltage detection circuit is repeated until the pre-charging of all N paths of loads is completed.

Description

Technical Field

[0001] This application relates to the field of power supply for embodied robots, and in particular to a power supply system and method for multiple alternating pre-charging. Background Technology

[0002] With the rapid development of embodied intelligence and humanoid robot technology, low-voltage DC power supply architecture has become the mainstream power supply architecture for humanoid robots. Humanoid robots have dozens of joint motors distributed throughout their bodies. To achieve refined power management and fault isolation, the joint motors are typically divided into multiple independent power supply paths based on their location, and each power supply path can be independently controlled for power on and off.

[0003] Joint motor drivers typically have large-capacity capacitors at their input terminals to stabilize the bus voltage and provide support for instantaneous high currents. However, at the moment of power-on, due to the equivalent short-circuit characteristics of the large-capacity capacitors, directly closing the main power supply switch will generate a huge inrush current, which may damage the switching devices, cause the bus voltage to drop, or trigger overcurrent protection. Therefore, before closing the main power supply switch, the load-side capacitor needs to be slowly charged through a pre-charging circuit, allowing the capacitor voltage to gradually rise to near the bus voltage, thereby suppressing the power-on inrush current.

[0004] In existing technologies, pre-charging schemes for multiple power supply paths either require each power supply path to be configured with an independent pre-charging resistor, with the number of pre-charging resistors being proportional to the number of power supply circuits. Since pre-charging resistors are usually power resistors, they are large in size, and multiple pre-charging resistors occupy a considerable area on the printed circuit board, which is not conducive to layout in space-constrained applications. At the same time, the large number of components leads to higher costs, dispersed heat dissipation paths are not conducive to centralized heat dissipation management, and there are more system failure points, resulting in lower reliability. Alternatively, the relays themselves are large in size, have slow switching speeds, limited lifespans, and are prone to contact arcing problems, making them unsuitable for applications with high requirements for size, reliability, and lifespan. More importantly, this centralized pre-charging scheme concentrates the pre-charging of all paths together, making it impossible to achieve independent power-on and power-off control for each power supply path. When a certain path needs to be powered off and then powered on separately, it is impossible to pre-charge that path independently without affecting the normal operation of other paths, lacking flexibility and fault isolation capabilities. Summary of the Invention

[0005] Based on this, it is necessary to address the problems in the existing technology, such as the large board area occupied by each independent pre-charging resistor, high cost, and the inability of centralized relay switching to independently power on and off and poor reliability. Therefore, a power supply system and method for multi-channel alternating pre-charging are proposed.

[0006] In a first aspect, a multi-channel alternating pre-charging power supply system is provided, comprising: Main power input terminal; A common pre-charge resistor, the first end of which is connected to the main power input terminal; An overcurrent protection device is connected in series between the main power input terminal and the first terminal of the common pre-charge resistor. The overcurrent protection device is an integrated electronic fuse chip. The fault status output terminal of the overcurrent protection device is connected to the control unit. N precharge branches, where N is an integer greater than 1. Each of the N precharge branches includes a precharge selection switch, which is a semiconductor switch. The first terminal of each precharge selection switch is connected to the second terminal of the common precharge resistor, and the second terminal of each precharge selection switch is connected to the power supply terminal of the corresponding load. N main power supply branches, each of the N main power supply branches includes a main power supply switch, the first end of the main power supply switch is connected to the main power input terminal, and the second end of the main power supply switch is connected to the power supply terminal of the corresponding load; N-channel voltage detection circuits, wherein the detection terminal of each of the N-channel voltage detection circuits is connected to the power supply terminal of the corresponding load, for detecting the voltage on the load side; and The control unit is electrically connected to the control terminals of each of the precharge selection switches, the control terminals of each of the main power supply switches, and the output terminals of each of the voltage detection circuits. The control unit is configured to: sequentially select each of the precharge selection switches in a preset order, so that the common precharge resistor limits the current charging of each load-side capacitor at different time periods, and at most only one of the precharge selection switches is in the on state at any given time; when a voltage detection circuit of the N voltage detection circuits detects that the load-side voltage of that circuit reaches a preset threshold, close the main power supply switch of that circuit and open the precharge selection switch of that circuit, so that the common precharge resistor is disconnected from the precharge branch of that circuit; when the continuous on time of a precharge selection switch exceeds the preset maximum precharge time and the load-side voltage of that circuit does not reach the preset threshold, open the precharge selection switch of that circuit, stop the precharge of that circuit, and do not close the main power supply switch of that circuit.

[0007] Secondly, a multi-channel alternating pre-charging method is provided, applied to the aforementioned multi-channel alternating pre-charging power supply system. The power supply system includes a main power input terminal, a common pre-charging resistor, N pre-charging branches, N main power supply branches, and N voltage detection circuits, where N is an integer greater than 1. The first terminal of the common pre-charging resistor is connected to the main power input terminal. Each pre-charging branch includes a pre-charging selection switch, which is a semiconductor switch. The first terminals of all pre-charging selection switches are commonly connected to the second terminal of the common pre-charging resistor. The second terminals of each pre-charging selection switch are respectively connected to the power supply terminal of a corresponding load. Each main power supply branch includes a main power supply switch. The first terminal of the main power supply switch is connected to the main power input terminal, and the second terminal of the main power supply switch is connected to the power supply terminal of a corresponding load. The detection terminal of each voltage detection circuit is connected to the power supply terminal of a corresponding load. The method includes: Step S1: System initialization, control all the precharge selection switches and all the main power supply switches to be in the off state; Step S2: Select the first precharge selection switch in a preset order so that the common precharge resistor limits the current of the load-side capacitor of that channel, and at most only one precharge selection switch is in the on state at any given time. Step S3: The voltage detection circuit of this circuit detects the load side voltage in real time. When the load side voltage reaches the preset threshold, the main power supply switch of this circuit is closed and the precharge selection switch of this circuit is opened to release the common precharge resistor. Step S4: Select the next precharge selection switch according to the preset order, and repeat step S3 until all N loads have completed precharging.

[0008] As can be seen from the technical solution provided in this application: On the one hand, by setting a common pre-charge resistor and connecting the first terminals of the pre-charge selection switches in the N pre-charge branches to the second terminal of the common pre-charge resistor, multiple power supply paths share the same pre-charge resistor. Compared with the prior art, this reduces the number of pre-charge resistors from equal to the number of power supply circuits to one, thereby reducing the area occupied by power resistors on the printed circuit board. This is beneficial for layout in space-constrained application scenarios. At the same time, the reduction in the number of components helps to reduce material costs, simplify heat dissipation design, and reduce system failure points, thereby improving system reliability. On the other hand, the pre-charge selection switch uses a semiconductor switch, which has the advantages of small size, fast switching speed, long service life, and no contact arc. It can adapt to application scenarios with space constraints, high reliability, and long service life requirements. Each power supply path is equipped with an independent pre-charge selection switch, main power supply switch, and voltage detection circuit. The control unit can independently control the pre-charge and switch the main power supply for each path, realizing independent power-on and power-off control for each power supply path. When a certain path needs to be powered off and then powered on again, it can be pre-charged separately without affecting the power supply. The control unit is configured to sequentially select each precharge selection switch in a preset order, allowing the common precharge resistor to perform current-limited charging of the capacitors on each load side at different time periods. At most, only one precharge selection switch can be active at any given time. Because the precharge process is instantaneous, the precharge times of each circuit can be completely staggered. A single common precharge resistor can meet the precharge requirements of multiple circuits through time-division multiplexing, reducing the number of precharge resistors while ensuring the precharge effect. The control unit is configured to... When a voltage detection circuit detects that the load-side voltage of a certain path reaches a preset threshold, it closes the main power supply switch for that path and opens the precharge selection switch to release the common precharge resistor. Based on the actual detected load-side voltage, it determines that precharging is complete and switches the main power supply switch accordingly. This ensures that each load capacitor is fully precharged before the main power supply switch closes, effectively suppressing inrush current when the main power supply switch closes. After precharging is complete, the precharge selection switch is promptly opened to release the common precharge resistor, allowing it to be immediately used for precharging the next path, thus improving overall precharging efficiency. In summary, the technical solution of this application reduces device area by using a single common resistor for time-division multiplexing, and enables independent power-on and power-off of semiconductor switches for each path's precharging, ensuring safety and reliability. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein: Figure 1 A schematic diagram of a power supply system structure for multiple alternating pre-charging provided in an embodiment of this application; Figure 2 This is a schematic diagram of the connection topology of the overcurrent protection device in the power supply system provided in the embodiments of this application; Figure 3 A schematic diagram illustrating the characteristics of the two-stage overcurrent protection provided in this application embodiment; Figure 4 A system block diagram of a multi-channel alternating pre-charge power supply system provided in this application embodiment applied to an embodied intelligent humanoid robot. Detailed Implementation

[0010] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0011] Joint motor drivers typically have large-capacity capacitors at their input terminals to stabilize the bus voltage and provide instantaneous high current support. However, at power-on, due to the equivalent short-circuit characteristics of these large-capacity capacitors, directly closing the main power supply switch can generate a huge inrush current, potentially damaging switching devices, causing a bus voltage drop, or triggering overcurrent protection. Therefore, before closing the main power supply switch, a pre-charging circuit is needed to slowly charge the load-side capacitor, gradually raising its voltage to near the bus voltage, thereby suppressing the power-on inrush current. In existing technologies, there are two main pre-charging schemes for multiple power supply paths: The first scheme uses an independent pre-charging circuit for each path. Each power supply path is equipped with an independent pre-charging circuit, consisting of a pre-charging resistor and a pre-charging switch connected in series and in parallel across the main power supply switch for that path. Upon power-on, the pre-charging switch is closed first, and the current is limited by the pre-charging resistor to slowly charge the load capacitor for that path. When the capacitor voltage rises to near the bus voltage, the main power supply switch is then closed, completing the pre-charging process. However, in this solution, each power supply path requires an independent pre-charging resistor, and the number of pre-charging resistors is proportional to the number of power supply circuits. Since pre-charging resistors are typically power resistors, they are relatively large, and multiple pre-charging resistors occupy a considerable area on the printed circuit board, making them unsuitable for space-constrained applications. Furthermore, the large number of components leads to higher costs, dispersed heat dissipation paths hinder centralized heat management, and numerous system failure points result in lower reliability. The second solution is centralized pre-charging with relay switching, where a single pre-charging resistor is switched to different power supply paths via a relay. However, relays themselves are large, have slow switching speeds, limited lifespans, and are prone to contact arcing, making them unsuitable for applications with high requirements for size, reliability, and lifespan. More importantly, this centralized pre-charging solution concentrates pre-charging for all paths, making independent power-on / off control for each power supply path impossible—when a path needs to be powered off and then re-powered, it cannot be pre-charged independently without affecting the normal operation of other paths, lacking flexibility and fault isolation capabilities.

[0012] Figure 1 This is a schematic diagram of the structure of a multi-channel alternating pre-charging power supply system provided in an embodiment of this application.

[0013] Figure 2 This diagram illustrates the connection topology of the overcurrent protection device in a power supply system, as provided in this embodiment. The diagram shows the connection method of the overcurrent protection device (integrated electronic fuse chip, eFuse) connected in series between the main power input terminal and the first terminal of the common pre-charge resistor, as well as the signal path connecting the fault status output terminal (FAULT pin) of the overcurrent protection device to the control unit. The diagram clearly shows that the overcurrent protection device is located on the common path of all downstream power supply paths, and a single device can achieve total overcurrent protection for the entire power supply system.

[0014] Figure 3 This diagram illustrates the characteristics of the two-stage overcurrent protection provided in this application embodiment. With time on the horizontal axis and current on the vertical axis, the diagram shows the relative thresholds of the first-stage short-circuit protection and the second-stage continuous overcurrent protection, as well as their respective response time magnitudes. As can be seen in the diagram, the overcurrent threshold of the first-stage short-circuit protection is much higher than the total current of all loads operating normally, with a response time in the microsecond range, triggering only in the event of a severe fault such as a short circuit. The overcurrent threshold of the second-stage continuous overcurrent protection is higher than the total current of all loads operating normally but lower than the overcurrent threshold of the first stage, with a response time in the millisecond range, allowing for short-term overloads (such as transient motor startup) but protecting against continuous overloads.

[0015] Figure 4 The diagram illustrates a system block diagram of a multi-path alternating pre-charging power supply system provided in this application for an embodied intelligent humanoid robot. The diagram shows the overall architecture of a 48V DC power supply passing through multiple alternating pre-charging circuits to power five loads: the head joint motor, left arm joint motor, right arm joint motor, left leg joint motor, and right leg joint motor. This demonstrates the specific application of this application in a five-path power supply scenario for a humanoid robot.

[0016] To address the aforementioned problems in the prior art, this application proposes a multi-channel alternating pre-charging power supply system. (Refer to...) Figure 1 The power supply system includes a main power input terminal, a common pre-charge resistor, N pre-charge branches, N main power supply branches, N voltage detection circuits, and a control unit, where N is an integer greater than 1. For ease of explanation, Figure 1 The example shown uses N equal to 5, but those skilled in the art should understand that N can take any integer greater than 1, and this application does not limit this.

[0017] The main power input terminal is used to connect an external DC power source to provide input voltage for the entire power supply system. In this embodiment, the DC voltage input to the main power input terminal can be 48V, but it is not limited to this and can also be other suitable DC voltage values.

[0018] The first terminal of the common pre-charge resistor is connected to the main power input terminal. This common pre-charge resistor is one of the core components of this application; its function is to limit the charging current during pre-charging, preventing a large surge current from being generated by the large-capacity capacitor on the load side due to an equivalent short circuit at power-on. Unlike existing technologies where each power supply path is equipped with an independent pre-charge resistor, this application uses only one common pre-charge resistor, shared by N power supply paths, thereby significantly reducing the number of power resistors.

[0019] Each of the N precharge branches includes a precharge selection switch, which is a semiconductor switch. The first terminal of each precharge selection switch is connected to the second terminal of a common precharge resistor, and the second terminal of each precharge selection switch is connected to the power supply terminal of its corresponding load. The function of the precharge selection switch is to select the precharge path for its corresponding route, allowing the common precharge resistor to be connected to the capacitor on the load side of that route for current-limited charging. Since the first terminals of each precharge selection switch are connected to the second terminal of the common precharge resistor, when a precharge selection switch is turned on, the common precharge resistor is connected to that precharge path; when the precharge selection switch is turned off, the common precharge resistor is released from that precharge path and can be used by other routes. Semiconductor switches are used instead of relays for the precharge selection switches because semiconductor switches have advantages such as small size, fast switching speed, long service life, and no contact arcing, making them suitable for applications with limited space, high reliability, and long lifespan requirements.

[0020] Each of the N main power supply branches includes a main power supply switch. The first terminal of the main power supply switch is connected to the main power input terminal, and the second terminal is connected to the power supply terminal of the corresponding load. The main power supply switch is the main power switch for each power supply path, closing after pre-charging to provide a direct, low-impedance power supply path to the load. The first terminal of the main power supply switch is directly connected to the main power input terminal, rather than the second terminal of the common pre-charging resistor. This means that under normal power supply conditions, the current does not pass through the common pre-charging resistor, thus avoiding continuous power loss and heat generation during normal operation.

[0021] Each of the N voltage detection circuits has its detection terminal connected to the power supply terminal of its corresponding load to detect the voltage on that load side. The voltage detection circuit is crucial for achieving adaptive pre-charge completion determination. It monitors the voltage of the capacitor on each load side in real time and sends the detection results to the control unit. By determining pre-charge completion based on the actual detected capacitor voltage rather than a fixed time, it can adapt to different load capacitor values, ensuring each circuit is fully pre-charged while avoiding unnecessary waiting time.

[0022] The control unit is electrically connected to the control terminals of each precharge selection switch, each main power supply switch, and the output terminals of each voltage detection circuit. The control unit is the core of the entire power supply system, responsible for controlling the on / off sequence of each switch based on preset logic and voltage detection results. In this embodiment, the control unit can be a microcontroller (MCU), but is not limited to it; it can also be other devices with logic control and signal acquisition capabilities, such as a digital signal processor (DSP) or a field-programmable gate array (FPGA).

[0023] The control unit is configured to sequentially select each precharge selection switch in a preset order, so that the common precharge resistor limits the current charging of each load-side capacitor at different time periods, and at most only one precharge selection switch is in the on state at any given time; when a voltage detection circuit of the N voltage detection circuit detects that the voltage of the load side reaches a preset threshold, it closes the main power supply switch of that circuit and opens the precharge selection switch of that circuit, so that the common precharge resistor is disconnected from the precharge branch of that circuit.

[0024] The aforementioned control logic is the core of this application's implementation of time-division multiplexing of a single common pre-charging resistor. Specifically, since the pre-charging process is instantaneous (typically requiring only a few hundred milliseconds), and the pre-charging times of each path can be completely staggered, a single common pre-charging resistor can sequentially select each path's pre-charging selection switch to simultaneously limit the current charging of each load-side capacitor at different time periods. The control unit ensures that at most one pre-charging selection switch is in the on state at any given time, thereby avoiding the problem of overloading the common pre-charging resistor or mutual interference of pre-charging currents caused by multiple paths pre-charging simultaneously. When the voltage on a certain load side reaches a preset threshold, the control unit determines that the pre-charging of that path is complete, then closes the main power supply switch for that path to bring it into normal power supply mode, and then opens the pre-charging selection switch for that path, so that the common pre-charging resistor is disconnected from the pre-charging branch of that path, allowing the common pre-charging resistor to be immediately used for the pre-charging of the next path. This process is repeated alternately until all loads have completed pre-charging.

[0025] Through the above architecture, this application reduces the number of pre-charge resistors from equal to the number of power supply circuits to a single resistor, thereby reducing the area occupied by power resistors on the printed circuit board and facilitating layout in space-constrained applications. Simultaneously, the reduction in the number of components helps lower material costs, simplify heat dissipation design (heat is concentrated in a single resistor, facilitating unified heat dissipation management), and reduce system failure points, thus improving system reliability. Furthermore, each circuit is equipped with an independent pre-charge selection switch, main power supply switch, and voltage detection circuit, allowing each circuit to independently switch between pre-charge and main power supply. When a circuit needs to be powered down and then powered back on separately, pre-charging can be performed on that circuit alone without affecting the normal operation of other circuits, providing excellent flexibility and fault isolation capabilities.

[0026] Based on the aforementioned overall system architecture, in order to ensure that the common pre-charge resistor can operate stably and reliably under the time-division multiplexing architecture of this application, Figure 1The common pre-charge resistor can be a surface-mount planar non-inductive power resistor. Its resistance value is configured to limit the initial peak pre-charge current within a preset safe range, and its rated power is configured to withstand energy loss during the pre-charge process. Specifically, the selection of a surface-mount planar non-inductive power resistor is mainly based on the following considerations: First, the surface-mount package has a small size, which can effectively save the layout space of the printed circuit board, which is especially important for applications with extremely limited internal space, such as humanoid robots. Second, the planar structure of the power resistor has a large heat dissipation area, which is conducive to the heat conduction to the outside through the copper foil of the printed circuit board, thereby improving heat dissipation conditions. Third, the non-inductive design ensures that the resistor does not generate significant parasitic inductive electromotive force during the transient process of rapid switching of the pre-charge selection switch, avoiding the impact of voltage spikes caused by parasitic inductance on the switching devices and load capacitance, and improving the stability of circuit operation.

[0027] Regarding the configuration of the common pre-charge resistor value in the embodiments of this application, the basic principle is that the load-side capacitor voltage is zero at the initial moment of pre-charge. At this time, the current of the pre-charge circuit is determined only by the main power input voltage and the resistance value of the common pre-charge resistor, that is, the initial peak current is approximately equal to the input voltage divided by the pre-charge resistor value. Therefore, by reasonably configuring the resistance value of the common pre-charge resistor, the initial peak current of pre-charge can be limited within a preset safe range, avoiding both excessive current causing impact on the upstream power supply and switching devices, and excessively low current leading to excessively long pre-charge time. As an embodiment of this application, when the main power input voltage is 48V and the single-channel load-side capacitor is approximately 2000 microfarads, if the expected pre-charge time is approximately 300 milliseconds, the resistance value of the common pre-charge resistor can be approximately 50 ohms. At this time, the initial peak current is approximately 0.96 amperes, which is within the safe range. Those skilled in the art should understand that the above values ​​are only examples, and the actual resistance value can be flexibly configured according to parameters such as input voltage, load capacitor value, and expected pre-charge time.

[0028] Regarding the configuration of the rated power of the common pre-charging resistor in this application embodiment, the total energy consumed by the common pre-charging resistor during the pre-charging process is equal to the electric field energy ultimately stored in the load-side capacitor. Therefore, the rated power of the common pre-charging resistor should be configured to withstand the aforementioned energy loss during the pre-charging process, while leaving an appropriate margin to ensure that the resistor will not be damaged due to overheating during the pre-charging process. Since the common pre-charging resistor in this application only conducts and consumes energy for a short period of time during the pre-charging process and is not connected to the main power supply path under normal power supply conditions, its rated power selection is mainly based on the instantaneous energy and average power during the pre-charging process, rather than the continuous power consumption. As an embodiment of this application, under the above parameter combination of 48V, 2000µF, and 50 ohms, the total energy consumed by the resistor during the pre-charging process is approximately 2.16 joules, and the average power during the pre-charging period is approximately 7.7 watts. A power resistor with a rated power of 15 watts can be selected to leave approximately twice the margin. Similarly, the above values ​​are only examples, and the actual rated power should be configured according to the specific application parameters.

[0029] In one embodiment of this application, both the precharge selection switch and the main power supply switch can be MOSFETs, i.e., metal-oxide-semiconductor field-effect transistors. The selection of MOSFETs as the precharge selection switch and main power supply switch is primarily based on the following advantages. First, MOSFETs are voltage-driven devices with simple gate drive circuits, low power consumption, and are easily driven directly by the control unit via general-purpose input / output pins (GPIO), simplifying the design of the control circuit. Second, MOSFETs have fast switching speeds, with turn-on and turn-off times typically on the order of nanoseconds to microseconds, meeting the requirements for rapid switching during precharge. Especially in scenarios with multiple alternating precharges, the precharge selection switch needs to frequently turn on and off to release and occupy the common precharge resistance; the fast switching characteristics help shorten the switching transition time and improve precharge efficiency. Third, MOSFETs have low on-resistance when turned on, resulting in low conduction losses for the main power supply switch under normal power supply conditions, which is beneficial for improving system efficiency and reducing heat generation. Finally, MOSFETs are contactless semiconductor devices, which do not have the contact wear, arcing, and lifespan limitations of mechanical relays. They have high reliability and long service life, making them suitable for applications with extremely high reliability requirements, such as humanoid robots.

[0030] In practical implementation, both the precharge selection switch and the main power supply switch can use N-channel enhancement-type MOSFETs. For the main power supply switch, since it needs to carry the continuous current of each load during normal operation, a MOSFET with low on-resistance and high rated current should be selected. For the precharge selection switch, since it only conducts for a short time during the precharge process, and the precharge current is limited by the common precharge resistance, its rated current requirement is relatively low, but its switching speed and withstand voltage rating should be considered. Those skilled in the art can select appropriate MOSFETs based on the voltage, current, and switching frequency requirements of the actual application; this application does not impose any limitations on this.

[0031] It should be noted that in the above embodiments, the preset threshold used to close the main power supply switch of a certain voltage detection circuit can be a preset proportion of the input voltage at the main power supply input terminal. The reason for setting the preset threshold as a preset proportion of the input voltage at the main power supply input terminal, rather than a fixed absolute voltage value, is that the fundamental goal of pre-charging is to gradually increase the load-side capacitor voltage to near the bus voltage, so that when the main power supply switch is closed, the difference between the capacitor voltage and the bus voltage is small enough to limit the inrush current within a safe range. Therefore, the criterion for determining the completion of pre-charging should be linked to the bus voltage (i.e., the input voltage at the main power supply input terminal), rather than using a fixed threshold unrelated to the bus voltage. The advantage of using a proportional form is that when the main power supply input voltage fluctuates or different input voltages are used in different application scenarios, the preset threshold can adaptively follow the changes in the input voltage, always ensuring that the load-side capacitor voltage reaches the corresponding proportion of the input voltage when pre-charging is complete, thereby ensuring the consistency of the pre-charging effect.

[0032] In one embodiment of this application, the preset ratio can be any value between 90% and 98% of the input voltage, with a typical value of 95%. When the preset ratio is 95%, if the main power input voltage is 48V, the preset threshold is 45.6V. At this time, the difference between the load-side capacitor voltage and the bus voltage is approximately 2.4V, and the inrush current when closing the main power supply switch is within a safe range. The selection of the preset ratio needs to strike a balance between the pre-charging effect and the pre-charging time: the higher the ratio, the closer the capacitor voltage is to the bus voltage when pre-charging is complete, and the smaller the inrush current when closing the main power supply switch, but the longer the pre-charging time; the lower the ratio, the shorter the pre-charging time, but the larger the inrush current when closing the main power supply switch. Those skilled in the art can select a suitable preset ratio based on the system's tolerance to inrush current and the requirements for power-on time.

[0033] In practical circuits, the load-side voltage during pre-charging is not an ideally smooth rising curve, but may be superimposed with fluctuations such as switching noise, power supply ripple, and load transient disturbances. If the control unit determines that pre-charging is complete and closes the main power supply switch based solely on a single voltage sampling result reaching a preset threshold, a misjudgment will occur when the voltage momentarily reaches the threshold due to the aforementioned fluctuations, but the actual average voltage has not yet reached it. This will cause the main power supply switch to close before the capacitor is actually fully charged, resulting in an unexpected inrush current. More seriously, if the voltage fluctuates repeatedly around the threshold, it may also cause the control unit to repeatedly switch between "pre-charging complete" and "not complete," affecting the stability of the system.

[0034] Therefore, based on the aforementioned preset threshold using a preset ratio of input voltage, this application embodiment introduces a determination mechanism based on continuous multi-sampling cycles to further improve the reliability of pre-charging completion determination. Specifically, the control unit in the aforementioned multi-channel alternating pre-charging power supply system is configured to determine that pre-charging is complete and close the main power supply switch for a given channel only when the load-side voltage reaches the preset threshold for multiple consecutive sampling cycles. This embodiment requires the control unit to confirm that the load-side voltage has reached the preset threshold for multiple consecutive sampling cycles before outputting the pre-charging completion determination result and closing the main power supply switch. This mechanism essentially introduces time-domain filtering or hysteresis at the software level: pre-charging is considered truly complete only when the voltage remains consistently and stably above the threshold, effectively filtering out false judgments caused by instantaneous voltage fluctuations. The specific number of consecutive sampling cycles can be configured according to the sampling frequency of the voltage detection circuit and the actual noise level. Higher sampling frequencies and greater noise levels require a more appropriate increase in the number of consecutive sampling cycles. As a lower-level embodiment, when the sampling frequency of the voltage detection circuit is about 1 kHz, it is required that the threshold is reached for no less than 10 consecutive sampling cycles (i.e., about 10 milliseconds) before the determination is completed. This time window is sufficient to filter out most high-frequency noise and transient disturbances, without significantly increasing the total precharge time.

[0035] Through the aforementioned continuous multi-sampling cycle determination mechanism, this application significantly improves the reliability of pre-charge completion determination without increasing additional hardware costs, solely through the software logic of the control unit. This ensures that the main power supply switch is closed only after each load-side capacitor has been fully pre-charged, thereby limiting the inrush current when the main power supply switch is closed to a safe range and improving the safety and stability of the entire power supply system during power-on.

[0036] In a time-division multiplexing architecture using a single common precharge resistor, precharging must proceed sequentially, with the common precharge resistor being a critical resource shared by all circuits. Under normal circumstances, each circuit's precharging is completed within hundreds of milliseconds, after which the common precharge resistor is released for the next circuit. However, if a circuit experiences an abnormal fault—such as a short circuit or failure of the load-side capacitor, a malfunction in the precharge selection switch preventing normal conduction, or other abnormalities in the load—the load-side voltage may fail to rise to the preset threshold, causing the precharging process to continue indefinitely. If the control unit does not intervene, the common precharge resistor will remain occupied by the faulty circuit, preventing all subsequent normal circuits from precharging and paralyzing the entire power supply system. This problem is unique to the single-resistor time-division multiplexing architecture—in existing technologies with independent precharge circuits, precharging is independent, and a fault in one circuit does not hinder the precharging process of others. Therefore, based on the aforementioned overall system architecture and alternating precharge control logic, this application embodiment further introduces a precharge timeout protection mechanism to address possible abnormal situations under the single resistor time-division multiplexing architecture. That is, the control unit in the aforementioned multi-path alternating precharge power supply system is also configured to: when the continuous conduction time of a certain path precharge selection switch exceeds the preset maximum precharge time and the voltage on the load side of that path does not reach the preset threshold, disconnect the path precharge selection switch, stop the path precharge, and do not close the path main power supply switch.

[0037] Specifically, this application sets a preset maximum pre-charge time for each pre-charge path. The control unit starts timing simultaneously when a pre-charge selection switch for a particular path is closed. If the load-side voltage of that path fails to reach a preset threshold within the preset maximum pre-charge time, an anomaly is determined for that path, and the following protective actions are executed: the pre-charge selection switch for that path is disconnected, stopping the pre-charge process for that path; and the main power supply switch for that path is not closed to prevent the faulty load from being directly connected to the bus and causing the fault to escalate. Through these actions, the common pre-charge resistance is released in a timely manner, allowing continued pre-charge service for subsequent normal paths, thereby effectively isolating the faulty path from the normal path and ensuring the overall availability of the system under a single-path fault.

[0038] The preset maximum precharge time should be greater than the time required for normal precharge to avoid accidental interruption of the normal precharge process, while not being too large to ensure that faults can be detected and handled in a timely manner. As a lower-level embodiment, when the normal precharge time for a single channel is approximately 300 to 500 milliseconds, the preset maximum precharge time can be approximately 2 seconds. This value is approximately 4 to 6 times the normal precharge time, providing sufficient margin to accommodate differences in load capacitance and input voltage fluctuations, while also enabling protection actions to be completed within seconds of a fault occurring. Those skilled in the art can configure the preset maximum precharge time according to the normal precharge time and fault response speed requirements in actual applications.

[0039] By using the precharge timeout protection mechanism, this embodiment solves the unique problem of the faulty path "locking up" the common precharge resistor in the single resistor time-division multiplexing architecture without increasing the additional hardware cost, and only through the software logic of the control unit. This ensures that the system can still complete the precharge and power-on normally when some paths fail, which significantly improves the fault tolerance and reliability of the power supply system.

[0040] Based on the aforementioned overall system architecture, to further enhance the safety of the power supply system under overcurrent and short-circuit faults, this embodiment adds a total overcurrent protection device. (Refer to...) Figure 2 The figure shows the connection topology of the overcurrent protection device in the system.

[0041] The multi-path alternating pre-charging power supply system also includes an overcurrent protection device. This overcurrent protection device is connected in series between the main power input and the first terminal of the common pre-charging resistor. The overcurrent protection device is an integrated electronic fuse chip, and its fault status output is connected to the control unit. Connecting the overcurrent protection device in series between the main power input and the first terminal of the common pre-charging resistor means that the device is located on the common path of all downstream power supply paths—whether it's the pre-charging branch or the main power supply branch, the current passes through this overcurrent protection device. Therefore, only one overcurrent protection device is needed to achieve total overcurrent protection for the entire power supply system, eliminating the need for separate protection devices on each branch, thus saving on the number of components and printed circuit board area. This layout is consistent with the core architecture of the single-resistor time-division multiplexing in this application, both embodying the design concept of "shared protection resources for common paths."

[0042] The overcurrent protection device employs an integrated electronic fuse chip (eFuse). An integrated electronic fuse chip is a protection device that integrates a power switch, current sampling amplifier, comparator, and control logic into a single chip. Compared to traditional fuses or resettable fuses, it offers advantages such as programmable overcurrent thresholds, fast response speed, fault status output, and automatic retry or latch-up. In this embodiment, the eFuse integrates a power MOSFET connected in series in the main power supply path. When the detected current exceeds the set threshold, the internal control logic quickly turns off the power MOSFET, cutting off the power supply path and thus protecting downstream circuits from overcurrent or short-circuit damage.

[0043] The fault status output terminal of the overcurrent protection device is connected to the control unit. The fault status output terminal (usually labeled the FAULT pin) outputs a valid fault signal when the overcurrent protection device triggers protection. The control unit can detect this signal to determine that an overcurrent fault has occurred in the system. Connecting the fault status output terminal to the control unit is significant because it achieves synergy between hardware-level rapid protection and system-level fault management: when a serious fault such as a short circuit occurs, the eFuse automatically cuts off the power supply within microseconds, without the need for control unit intervention, thus ensuring timely protection; simultaneously, the eFuse notifies the control unit of the fault event via the FAULT pin. Upon receiving the fault signal, the control unit can perform corresponding system-level processing, such as disconnecting all precharge selection switches and the main power supply switch, reporting fault information to the host computer, recording the fault log, or initiating automatic retry when conditions permit, thereby realizing a complete closed loop from hardware protection to system management.

[0044] By setting up the above-mentioned total overcurrent protection device, this application adds a hardware safety barrier to the common path of the single resistor time-division multiplexing architecture. One device can protect all downstream paths, ensuring rapid disconnection under overcurrent and short-circuit faults, and realizing system-level fault management through the linkage between the fault status output terminal and the control unit, which significantly improves the safety and maintainability of the power supply system.

[0045] In a multi-rail motor power supply system, the total bus current dynamically changes with the operating status of each motor, resulting in complex current characteristics. Under normal operation, the total steady-state current of each motor is within a relatively stable range; however, during motor startup, acceleration, or sudden load changes, short-duration transient high currents are generated, with peak values ​​potentially significantly higher than the steady-state total current. Furthermore, in the event of severe faults such as bus short circuits or capacitor breakdowns, the current can surge to levels far exceeding any normal operating current within an extremely short time. Using a single overcurrent threshold and response time for protection presents a dilemma: if the threshold is set too low and the response too fast, it may falsely trigger protection during normal transient high currents such as motor startup, leading to abnormal power outages; if the threshold is set too high and the response too slow, it may fail to cut off power in time during short-circuit faults, potentially causing device damage or even fire.

[0046] To address the aforementioned issues, this embodiment employs a two-stage overcurrent protection strategy, setting differentiated thresholds and response times for overcurrent events of different natures. Specifically, based on the aforementioned overall overcurrent protection device, to ensure the reliability of fault protection while avoiding false triggering under normal operating conditions, this application further refines the protection strategy of the overcurrent protection device. (Refer to...) Figure 3The figure shows the characteristic curves of two-stage overcurrent protection. The overcurrent protection device is configured with two stages of overcurrent protection: the first stage is short-circuit protection, with an overcurrent threshold much higher than the total current of each load when operating normally, and a response time in the microsecond range; the second stage is continuous overcurrent protection, with an overcurrent threshold higher than the total current of each load when operating normally but lower than the overcurrent threshold of the first stage, and a response time in the millisecond range.

[0047] Specifically, the first-level short-circuit protection has an overcurrent threshold much higher than the total current of all loads operating normally, with a response time in the microsecond range. This level of protection is specifically designed for severe faults such as short circuits and capacitor breakdown. Since the current characteristic of these faults spikes to levels far exceeding normal operating current in an extremely short time, setting the threshold to be much higher than the normal total current ensures that this level of protection will not falsely trigger under any normal operating conditions (including motor starting transients). Simultaneously, the microsecond-level response time ensures that the power supply can be quickly cut off after a short-circuit fault occurs, limiting the duration of the fault current to a very short range and preventing damage to power devices and lines due to overcurrent and overheating. As a lower-level embodiment, when the typical total current of the five-rail motor is approximately 15 to 25 amps during normal operation, and the peak transient current may reach 30 to 35 amps, the overcurrent threshold of the first-level short-circuit protection can be set to approximately 100 to 150 amps, with a response time of less than 10 microseconds. This threshold is much higher than any normal operating current and will only trigger in the event of a true short-circuit fault.

[0048] The second level, continuous overcurrent protection, has an overcurrent threshold higher than the total current of all loads operating normally but lower than the first level's overcurrent threshold, with a response time in milliseconds. This level of protection is specifically designed for continuous overload faults, such as a stalled motor or multiple motors operating at high loads simultaneously for extended periods, resulting in a consistently high total current that has not yet reached a short-circuit level. Setting the threshold higher than the normal total current but lower than the first level threshold ensures that this level of protection does not trigger during normal steady-state operation but can operate under continuous overload conditions. The millisecond-level response time allows short-term transient large currents (e.g., the large current generated when a motor starts) to pass through without triggering the protection, responding only to overloads of longer duration, thus avoiding false interruptions to normal transient processes. As a subordinate embodiment, under the parameters of approximately 15 to 25 amps for normal total current and approximately 30 to 35 amps for peak transient current, the overcurrent threshold of the second-level continuous overcurrent protection can be set to approximately 35 to 45 amps (typical value 40 amps), and the response time is approximately 5 to 20 milliseconds (typical value 10 milliseconds). This threshold is higher than the normal peak transient current, and the millisecond-level response time is sufficient to skip short transient processes such as motor starting and only protect against continuous overload.

[0049] Through the aforementioned two-stage overcurrent protection strategy, this application achieves a fine balance between short-circuit protection and overload protection: the first stage, with a high threshold and microsecond-level response, is specifically designed to handle severe faults such as short circuits, ensuring rapid disconnection and preventing false triggering; the second stage, with a medium threshold and millisecond-level response, is specifically designed to handle sustained overloads, allowing normal transient large currents to pass through while only acting on sustained overloads. The two stages of protection work together to ensure safe protection under fault conditions while avoiding false triggering under normal operating conditions, significantly improving the protection reliability of the power supply system under complex dynamic loads.

[0050] Furthermore, in this embodiment of the power supply system for multi-channel alternating pre-charging, the control unit is configured to: during system power-on initialization, control all pre-charging selection switches and all main power supply switches to be in the off state. System power-on initialization is the starting point of the entire power supply system's workflow, and ensuring that all switches are in the off state at this stage is of significant safety importance. On the one hand, at the moment of power-on, the supply voltage of the control unit itself may not yet be stable, and the level state of its output pins may be in an uncertain state. If all switches are not actively configured to be in the off state, a switch may be accidentally turned on, causing the corresponding load to be directly connected to the bus without pre-charging, generating a huge surge current. On the other hand, initializing all switches to the off state lays a clean starting state for subsequent pre-charging in a preset order, avoiding pre-charging sequence disorder caused by the uncertainty of the switch state before power-on.

[0051] In its implementation, after power-on reset, the control unit first executes an initialization program, configuring all general-purpose input / output (GPIO) pins corresponding to the precharge selection switches and main power supply switches to output mode and outputting a shutdown level to ensure all switches are reliably disconnected. Subsequently, the control unit initializes peripherals such as the analog-to-digital converter (ADC), timers, and communication interfaces, entering a precharge-ready state, awaiting a precharge start command or automatic initiation of the precharge process. Through this initialization configuration, the system enters a safe state immediately after power-on, providing a reliable foundation for the subsequent alternating precharge process.

[0052] In a single-chip common precharge resistor-time-division multiplexing architecture, each load must be precharged sequentially, and the selection order of the precharge selection switch directly determines the power-on order of each load. In a fixed precharge sequence scheme, the power-on priority of each path is predetermined and cannot be changed, which may not meet actual needs in some scenarios. For example, in humanoid robot applications, during normal cold starts, it may be desirable to prioritize precharging the main control module and sensors in the head to enable them to work first and monitor and manage the subsequent power-on process; however, in some emergency situations (such as partial power-on after a system failure), it may be necessary to prioritize precharging specific joint motor branches to restore critical motion capabilities; furthermore, the power-on priority of each load may differ in different operating modes (such as debugging mode, normal operation mode, and low-power mode). If the precharge sequence is fixed and unconfigurable, it cannot flexibly adapt to the above diverse needs.

[0053] To address the aforementioned issues, based on the aforementioned alternating precharge control logic, this application further endows the control unit with the ability to dynamically configure the precharge sequence. This adapts to the differentiated power-on priority requirements of various loads in different application scenarios. Specifically, the control unit of the multi-channel alternating precharge power supply system is also configured to support the dynamic configuration of the selection order of each precharge selection switch, prioritizing the precharge of preset critical load branches. Specifically, the control unit maintains a precharge sequence configuration table, which records the selection order of each precharge selection switch. When executing the precharge process, the control unit sequentially selects each precharge selection switch according to the order recorded in the configuration table, thereby prioritizing the precharge of preset critical load branches. Critical load branches refer to load branches that require priority power-on in specific application scenarios, such as branches carrying core functions like the main control module and sensors, which have higher priority than ordinary joint motor drive branches. The precharge sequence configuration table can be dynamically updated during system operation, enabling real-time adjustment of the precharge sequence. By dynamically configuring the pre-charge sequence, this embodiment enables the power supply system to flexibly adjust the power-on order of various loads according to actual application scenarios and operating modes, prioritizing the timely power-on of critical loads and significantly improving the flexibility and adaptability of the power supply system. This capability is particularly important in systems such as humanoid robots that have diverse operating modes and complex power-on timing requirements.

[0054] Building upon the aforementioned dynamic configuration of the precharge sequence, this embodiment provides a specific configuration interface implementation method, enabling the host computer to conveniently issue precharge sequence configuration commands to the control unit. Specifically, in one embodiment of this application, the control unit is equipped with a Controller Area Network (CAN) communication interface. The control unit is configured to receive precharge sequence configuration commands from the host computer via the CAN communication interface and adjust the selection order of each precharge selection switch according to the configuration commands. In specific implementation, the control unit's CAN communication interface is connected to the system's CAN bus via a CAN transceiver. The host computer (typically the robot's main control module) sends the precharge sequence configuration command to the control unit via the CAN bus. This command can use a custom CAN data frame format, containing the identification information of each load and its corresponding precharge priority or selection order. After receiving the command, the control unit parses it, updates its internal precharge sequence configuration table according to the order specified in the command, and sequentially selects each precharge selection switch according to the updated configuration table in the subsequent precharge process. In this way, the host computer can adjust the precharge sequence at any time during system operation, achieving dynamic management of the power-on priority of each load.

[0055] In one embodiment of this application, the aforementioned precharge sequence configuration instruction may include a priority list, in which each channel is arranged from highest to lowest priority. After receiving the instruction, the control unit executes precharge according to the order of the list; alternatively, the instruction may assign a priority value to each channel, and the control unit sorts the channels from highest to lowest priority and then executes precharge. The specific instruction format and communication protocol can be defined by the system designer according to actual needs, and this application does not limit them.

[0056] Through the CAN communication interface, this embodiment provides a standardized and highly reliable way to dynamically configure the pre-charge sequence, enabling the host computer to easily remotely manage and dynamically adjust the power-on timing of the power supply system, and seamlessly integrate with the overall control architecture of the robot system.

[0057] The aforementioned ability to dynamically configure the pre-charge sequence typically relies on a host computer issuing configuration commands via a communication interface. However, a unique contradiction exists in system cold start scenarios: the host computer responsible for issuing the pre-charge sequence configuration commands (usually the robot's main control module) is itself one of the loads that needs to be pre-charged and powered on. During the cold start phase immediately after system power-on, the main control module has not yet completed power-on and initialization, and cannot issue any configuration commands to the power management control unit via the CAN bus or other communication interfaces. At this time, if the control unit simply waits for external pre-charge sequence configuration commands without initiating pre-charge, the entire system will fall into a deadlock state of "the main control module needs to configure the pre-charge sequence → but the main control module itself needs to be pre-charged and powered on first," the so-called "chicken and egg" problem. This problem is particularly prominent in distributed intelligent systems such as embodied humanoid robots where the main control module and the actuator share the same power supply system.

[0058] In view of this, based on the aforementioned dynamic configuration of the pre-charge sequence, this embodiment further provides a solution for the special scenario of system cold start, ensuring that the system can still autonomously complete pre-charge even without external configuration commands. Specifically, the control unit of the multi-channel alternating pre-charge power supply system is configured to: when the system is cold-started and no external pre-charge sequence configuration command is received, sequentially select each pre-charge selection switch according to a locally preset pre-charge sequence. The locally preset pre-charge sequence is: first pre-charge the load branch containing the control module and sensor, then pre-charge the load branches containing each joint motor in sequence. Specifically, this embodiment configures a locally preset pre-charge sequence for the control unit. When the system is cold-started and the control unit does not receive an external pre-charge sequence configuration command within a predetermined time, the control unit automatically switches to the locally preset mode and performs pre-charge according to the default pre-charge sequence stored in the control unit's non-volatile memory, without waiting for external commands. The locally preset pre-charge sequence is: first pre-charge the load branch containing the control module and sensor, then pre-charge the load branches containing each joint motor in sequence. This design sequence is logically sound: the control module (including the main control unit, power management unit, etc.) is responsible for the overall control and management of the system; sensors (including head cameras, wrist cameras, force sensors, etc.) are responsible for sensing the external environment and providing input for control decisions. Prioritizing the pre-charging of the control module and sensors, allowing them to operate before the joint motors, means that the system already possesses sensing and decision-making capabilities before the joint motors are powered on. This enables the system to monitor and manage the subsequent joint motor power-on process, promptly detecting and handling anomalies. Conversely, if the joint motors are powered on before the control module, the system is in a "brainless" state during the initial power-on phase, unable to effectively monitor the joint motors' operating status, posing a safety hazard. Therefore, setting the pre-charging priority of the control module and sensors higher than that of the joint motors is a reasonable design that aligns with the system's control logic and safety requirements.

[0059] In practical implementation, the locally preset pre-charge sequence can be fixed in the control unit's firmware or stored in external non-volatile memory (e.g., Flash, EEPROM), and read by the control unit during power-on initialization. After the main control module completes power-on initialization, it can send a new pre-charge sequence configuration command to the control unit via the CAN communication interface. Upon receiving the command, the control unit exits the locally preset mode and executes subsequent pre-charge operations according to the order specified by the host computer (e.g., in scenarios where some circuits are powered off and then powered on again). By combining the locally preset mode with external dynamic configuration, the system ensures both autonomous and reliable power-on during cold starts and retains flexible configuration capabilities during normal operation.

[0060] By designing a local preset pre-charge sequence for cold starts, this application cleverly solves the deadlock problem faced by dynamically configuring the pre-charge sequence in cold start scenarios under a single resistor time-division multiplexing architecture. This ensures that the system can autonomously, safely, and orderly complete the power-on process without external intervention. At the same time, the reasonable default pre-charge sequence ensures the priority power-on of the control module and sensors, laying the foundation for the safe operation of the system.

[0061] Based on the aforementioned general multi-path alternating pre-charge power supply system, this application specifically applies this technical solution to the power supply scenario of an embodied intelligent humanoid robot, so as to fully leverage the advantages of the single-resistor time-division multiplexing architecture in this specific scenario. (Refer to...) Figure 4 The figure illustrates the overall architecture of the five-way power supply system applied to the embodied intelligent humanoid robot of this application.

[0062] The power supply system is for the embodied intelligent humanoid robot. The main power input terminal receives 48V DC voltage, N equals 5, and the five loads correspond to the power supply paths of the joint motors of the five body parts of the humanoid robot.

[0063] An embodied humanoid robot is an intelligent robot system with a humanoid shape and structure, capable of perceiving its environment and autonomously performing physical tasks. Currently, 48V DC power supply has become the mainstream power supply architecture for humanoid robots. This is because the 48V voltage level achieves a good balance between power transmission efficiency, safety, and device maturity: compared to lower voltage levels (e.g., 12V, 24V), 48V draws less current and has lower line losses when transmitting the same power, which helps reduce cable weight and heat generation; compared to higher voltage levels, 48V is still within the safe extra-low voltage range, posing a lower risk of electric shock to humans, and the 48V power device (e.g., MOSFETs, capacitors, connectors, etc.) supply chain is mature and cost-controllable. Therefore, humanoid robots commonly use 48V DC power supply.

[0064] Humanoid robots have dozens of joint motors distributed throughout their bodies. To achieve refined power management and fault isolation, these joint motors are typically divided into multiple independent power supply paths based on body parts, each of which can be independently controlled for power on / off. In this embodiment, N equals 5, meaning the five load paths correspond to the power supply paths for the joint motors of the five body parts of the humanoid robot. Choosing five paths, rather than more or fewer, is based on a comprehensive consideration of the humanoid robot's structural characteristics and the time-division multiplexing architecture of a single resistor. Firstly, the humanoid robot's body structure can naturally be divided into five main parts: head, left arm, right arm, left leg, and right leg. Each part contains several joint motors. Grouping the joint motors of the same part into a single power supply path not only facilitates the structural layout but also enables independent power on / off and fault isolation by body part. On the other hand, considering the characteristics of the single-resistor time-division multiplexing architecture, the choice of the number of paths N needs to strike a balance between the benefits of resistor multiplexing and the total precharge time: if N is too small (e.g., 2 or 3 paths), the PCB area and cost savings from merging multiple precharge resistors into one are limited, and the advantages of single-resistor multiplexing are not significant; if N is too large (e.g., more than 10 paths), the total precharge time of each path will increase linearly, potentially exceeding the system's power-on timing requirements. Five paths strike this balance—reducing the number of precharge resistors from 5 to 1 significantly saves PCB area and BOM (Bill of Materials) costs, while keeping the total precharge time within an acceptable range for the system. Therefore, five paths are the optimal choice for matching the single-resistor time-division multiplexing architecture.

[0065] Under the specific parameters of a 48V five-channel power supply, the technical effects of this application are fully demonstrated. As one embodiment of this application, when the single-channel load-side capacitor is approximately 2000 microfarads and the common pre-charge resistor value is approximately 50 ohms, the single-channel pre-charge time is approximately 300 to 500 milliseconds, and the total pre-charge time for all five channels is approximately 1.5 to 2.5 seconds. This total pre-charge time fully meets the power-on sequence requirements of the humanoid robot. Simultaneously, the number of pre-charge resistors is reduced from five to one, and the PCB area of ​​the pre-charge region is reduced by approximately 70% to 80%, greatly alleviating the problem of limited space on the humanoid robot's power board.

[0066] By specifically applying the multi-path alternating pre-charge technology to the power supply scenario of five joint motors for a 48V humanoid robot, this embodiment achieves the optimal match between the technical solution and the application scenario. It fully leverages the advantages of the single resistor time-division multiplexing architecture in reducing components, shrinking PCB area, and lowering costs, while keeping the total pre-charge time within an acceptable range for the system. This effectively solves the practical engineering problems of limited power board space and high BOM cost for humanoid robots.

[0067] In one embodiment of this application, the five load paths are respectively the power supply paths for the head joint motor, the left arm joint motor, the right arm joint motor, the left leg joint motor, and the right leg joint motor. These five paths correspond one-to-one with the body structure of the humanoid robot, possessing clear physical meaning and engineering rationality. The head joint motor power supply path supplies power to the joint motors of the head (such as the neck pitch and yaw joints), as well as the head's main control module and sensors (such as cameras, LiDAR, microphone arrays, etc.); the left and right arm joint motor power supply paths supply power to the joint motors of the left and right arms (such as the shoulder, elbow, and wrist joints); and the left and right leg joint motor power supply paths supply power to the joint motors of the left and right legs (such as the hip, knee, and ankle joints).

[0068] Grouping joint motors of the same body part into a single power supply path has the following advantages: First, the joint motors of the same part are physically concentrated, making the routing and layout of the power supply line easier and reducing long-distance power cables across parts. Second, dividing by part enables regional isolation of faults. When a power supply path for a certain part fails, the power supply to that part can be disconnected without affecting the normal operation of other parts, improving the fault tolerance of the system. Third, independent power supply and shutdown for each part facilitates refined power management. For example, when the robot is in standby mode, only the head main controller and sensors can be powered while the power supply to the limb motors can be disconnected to reduce power consumption. When a specific action needs to be performed, the corresponding part can be powered on as needed.

[0069] In practical implementation, the load-side capacitor of each power supply path corresponds to the equivalent sum of the input capacitors of all joint motor drivers in that section. Since the number and power rating of joint motors in different sections may vary, the equivalent capacitance values ​​of the load-side capacitors may differ. This application employs a pre-charge completion determination mechanism based on actual voltage detection (rather than a fixed time), which can adapt to loads with different capacitance values, ensuring that each path is fully pre-charged before switching to the main power supply. Therefore, differences in capacitance values ​​between paths will not affect the consistency of the pre-charge effect.

[0070] In the above embodiments, each load-side capacitor can be a large-capacity capacitor, whose capacitance is configured to stabilize the bus voltage corresponding to the load and provide instantaneous high current support. The load-side capacitor is a key component at the input of the articulated motor driver, typically using large-capacity electrolytic capacitors or film capacitors. It primarily performs two functions in the power supply system. Firstly, it stabilizes the bus voltage corresponding to the load: During operation, especially during acceleration, deceleration, and sudden load changes, the articulated motor draws rapidly changing pulse currents from the bus. Without a large-capacity capacitor for buffering, these pulse currents will cause significant voltage drops across the parasitic inductance and resistance of the power supply line, leading to unstable bus voltage, affecting the normal operation of the motor driver, and even triggering undervoltage protection. The large-capacity capacitor, with its energy storage function, can release charge when current demand increases instantaneously and absorb charge when current demand decreases, thereby suppressing bus voltage fluctuations within a small range and ensuring power quality. Secondly, it provides instantaneous high current support: The articulated motor requires a large instantaneous current during startup or acceleration. If it relies entirely on the upstream power supply through long-distance cables, the parasitic impedance of the lines will limit the rate of current rise and generate a large voltage drop. A large-capacity capacitor is placed near the input terminal of the motor driver, which can provide a large instantaneous current nearby to meet the dynamic response requirements of the motor, while reducing the transient current burden on the upstream power supply and power supply lines.

[0071] Because the load-side capacitor has a large capacitance, it is equivalent to a short circuit at the moment of power-on. If the main power supply switch is closed directly to connect it to the bus, a huge inrush current will be generated. This is the fundamental reason for the need for a pre-charging circuit—by limiting the current through a pre-charging resistor, the voltage of the large-capacity capacitor is gradually increased to near the bus voltage, thereby limiting the inrush current to a safe range when the main power supply switch is closed. Therefore, the existence of the large-capacity capacitor on the load side and the necessity of the pre-charging circuit are inseparable: the larger the capacitance, the larger the inrush current when directly powered on, and the more urgent the need for a pre-charging circuit. The single-resistor time-division multiplexing pre-charging scheme of this application is designed for the power-on scenario of multiple large-capacity capacitor loads, and can significantly reduce the number of pre-charging resistors while ensuring the pre-charging effect.

[0072] In specific implementation, the capacitance value of the load-side capacitor is configured according to the number, power level and dynamic response requirements of the joint motors in the corresponding part. For example, as an embodiment of this application, the equivalent capacitance of the load side of a single joint motor is about 2000 microfarads. However, those skilled in the art should understand that the actual capacitance value can be flexibly selected according to the specific application, and this application does not limit it.

[0073] Considering that the number of power supply circuits may vary depending on the product model, configuration level, or application scenario in practical applications of multi-power supply systems, for example, the same series of humanoid robots may have a basic version (five power supply circuits) and a high-end version (more power supply circuits, such as adding independent power supply paths for the torso and hands). If the pre-charging circuit adopts a non-modular integrated design, the core circuit of the power board needs to be redesigned for each additional power supply circuit, including rearranging the connection relationships of the common pre-charging resistor, pre-charging selection switch, voltage detection circuit, and control unit. This not only increases design costs and cycle time but also hinders rapid product iteration and serialization. Therefore, to further improve the scalability of the power supply system, this application further optimizes the layout of the pre-charging branch and voltage detection circuit. That is, each pre-charging selection switch and the corresponding voltage detection circuit are integrated into an independent pre-charging functional module. The power supply system is configured such that when increasing the number of power supply circuits, only the pre-charging functional module needs to be added and connected to the second end of the common pre-charging resistor and the control unit, without replacing the common pre-charging resistor.

[0074] Specifically, this application adopts a modular design, integrating each precharge selection switch and its corresponding voltage detection circuit into an independent precharge function module. Each precharge function module is a self-contained functional unit, internally containing the precharge selection switch and its driving circuit, voltage detection circuit (such as a resistor divider network and an ADC interface), and providing standard connection interfaces: one end connects to the second terminal of the common precharge resistor (precharge current input), the other end connects to the power supply terminal of the corresponding load (precharge current output and voltage detection point), and there is a control interface connecting to the control unit (switch control signal and voltage detection signal). Since the interface standard of all precharge function modules is unified, when increasing the number of power supply circuits, it is only necessary to connect the newly added precharge function module in parallel to the second terminal of the common precharge resistor and the corresponding interface of the control unit, without modifying the core architecture.

[0075] Most importantly, there's no need to replace the common pre-charging resistor when increasing the number of power supply circuits. This is because the selection of the common pre-charging resistor is primarily based on the current and energy requirements of a single pre-charge circuit, not the number of power supply circuits. In a time-division multiplexing architecture, only one circuit uses the common pre-charging resistor at any given time. Therefore, regardless of the total number of circuits, the instantaneous current and energy per pre-charge are the same. Increasing the number of circuits only increases the total number of pre-charges and the total pre-charge time, without changing the operating stress of individual resistors. Therefore, the common pre-charging resistor can be used in configurations with different numbers of circuits without needing to be replaced. This characteristic makes modular design highly flexible and economical.

[0076] Through modular and scalable design, this application achieves serialized and platform-based development of the power supply system: the core common pre-charge resistor and control unit architecture remain unchanged, and the application requirements of different circuits can be adapted by adding or removing the number of pre-charge function modules. This shortens the product development cycle, reduces design costs, and improves the reusability and maintainability of the hardware platform. At the same time, the modular design also facilitates production testing and after-sales maintenance—when a pre-charge function module of a certain circuit fails, the module can be replaced individually without replacing the entire power board, reducing maintenance costs.

[0077] This application also provides a multi-channel alternating pre-charging method, which is applied to the power supply system of the multi-channel alternating pre-charging described in the foregoing embodiments. The power supply system includes a main power input terminal, a common pre-charging resistor, N pre-charging branches, N main power supply branches, N voltage detection circuits, and a control unit, where N is an integer greater than 1. The first terminal of the common pre-charging resistor is connected to the main power input terminal. Each pre-charging branch includes a pre-charging selection switch, which is a semiconductor switch. The first terminals of all pre-charging selection switches are connected to the second terminal of the common pre-charging resistor, and the second terminals of each pre-charging selection switch are respectively connected to the power supply terminal of a corresponding load. Each main power supply branch includes a main power supply switch, the first terminal of which is connected to the main power input terminal, and the second terminal of which is connected to the power supply terminal of a corresponding load. The detection terminal of each voltage detection circuit is connected to the power supply terminal of a corresponding load. The control unit is electrically connected to the control terminals of each pre-charging selection switch, each main power supply switch, and the output terminal of each voltage detection circuit. The multi-channel alternating pre-charging method includes the following steps S1 to S4, which are described in detail below: Step S1: System initialization, control all precharge selection switches and all main power supply switches to be in the off state.

[0078] Upon system power-up, the control unit first executes an initialization program, configuring the control pins corresponding to all precharge selection switches and all main power supply switches to output mode and outputting a shutdown level to ensure all switches are reliably disconnected. This step aims to establish a safe and deterministic starting state for the subsequent precharge process, preventing accidental switching on of a switch due to uncertain control pin levels at power-up, which could lead to inrush current as the load is directly connected to the bus without precharge. After completing switch initialization, the control unit also initializes peripherals such as the analog-to-digital converter (ADC), timers, and communication interfaces, and reads the precharge sequence configuration (locally preset or externally sent). It then enters the precharge-ready state, waiting to initiate the precharge process.

[0079] Step S2: Select the first precharge selection switch in the preset order so that the common precharge resistor limits the current charging of the load-side capacitor of that channel, and at most only one precharge selection switch is in the on state at the same time.

[0080] The control unit selects the first precharge selection switch according to the preset precharge sequence. At this time, current flows from the main power input through the common precharge resistor and the selected precharge selection switch to the load-side capacitor of that path. The common precharge resistor limits the charging current, causing the load-side capacitor voltage to rise exponentially. During this process, the control unit ensures that at most one precharge selection switch is in the on state at any given time, thus guaranteeing that the common precharge resistor serves only one load at any given moment, preventing overload of the common precharge resistor or mutual interference of precharge currents caused by multiple simultaneous precharges. Because the precharge selection switch uses a semiconductor switch, its turn-on and turn-off speeds are fast, allowing for rapid switching after the control unit issues a command, ensuring the accuracy of the precharge timing.

[0081] Step S3: The voltage on the load side of the circuit is detected in real time by the voltage detection circuit. When the voltage on the load side of the circuit reaches the preset threshold, the main power supply switch of the circuit is closed and the precharge selection switch of the circuit is opened so that the common precharge resistor is disconnected from the precharge branch of the circuit.

[0082] Step S3, the voltage detection and main power supply switching step, is the core of the entire pre-charge process. While the pre-charge selection switch for this path is on, the control unit collects the load-side voltage in real time at a fixed sampling frequency through the voltage detection circuit. The voltage detection circuit typically consists of a resistor divider network and an analog-to-digital converter, converting the high voltage on the load side into a low voltage signal that the control unit can collect. The control unit judges the collected voltage signal, and when the load-side voltage reaches a preset threshold, it determines that pre-charging for this path is complete. The preset threshold is usually set as a preset proportion of the main power input voltage, for example, 95% of the input voltage, to ensure that the load-side capacitor voltage is sufficiently close to the bus voltage. After determining that pre-charging is complete, the control unit performs two actions: first, it closes the main power supply switch for this path, allowing it to enter normal power supply mode. At this time, because the load-side capacitor voltage is close to the bus voltage, the inrush current when the main power supply switch is closed is extremely small and within a safe range; then, it opens the pre-charge selection switch for this path, releasing the common pre-charge resistor from this pre-charge path so that it can be used for the next pre-charge path. The timing design of closing the main power supply switch first and then opening the precharge selection switch ensures the continuity of power supply to the load and avoids power interruption at the moment of switching.

[0083] Step S4: Select the next precharge selection switch in the preset order, and repeat step S3 until all N loads have completed precharging.

[0084] After the first pre-charge is completed and the common pre-charge resistor is released, the control unit selects the next pre-charge selection switch according to a preset sequence, repeating the voltage detection and switching process in step S3. This cycle iterates, sequentially for the second, third, and so on, until the Nth channel completes pre-charge. When all N loads have completed pre-charge and entered normal power supply mode, all pre-charge selection switches remain open, all main power supply switches remain closed, the control unit reports the pre-charge completion status to the host computer, and the system enters normal operating mode.

[0085] Through the orderly execution of the above four steps, this method achieves time-division multiplexing precharging of multiple loads using a single common precharging resistor. The core of the entire process lies in the sequential selection—voltage determination—main switch switching—resistor release—next-path loop logic. This logic ensures that each load is fully precharged before being connected to the main power supply, effectively suppressing inrush current. Furthermore, by using time-division multiplexing of the common precharging resistor, the number of precharging resistors is reduced from N to 1, significantly saving PCB area and BOM cost. Simultaneously, since the precharging completion determination for each path is based on actual voltage detection rather than a fixed time, this method can adapt to different load capacitance values, ensuring the consistency of precharging effects across all paths.

[0086] In practical applications, the above method can also be combined with the aforementioned embodiments or their additional functions. For example, a continuous multi-sampling cycle determination mechanism can be introduced in step S3 to improve reliability; pre-charge timeout protection can be introduced in steps S2 and S3 to deal with faulty circuits; dynamic configuration of the pre-charge sequence can be introduced before step S2 to adapt to different scenarios; and two-level overcurrent protection can be introduced in the system common path to ensure safety, etc. These additional functions work together with the basic method to form a complete, safe and reliable multi-path alternating pre-charge scheme.

[0087] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A power supply system with multiple alternating pre-charging channels, characterized in that, include: Main power input terminal; A common pre-charge resistor, the first end of which is connected to the main power input terminal; An overcurrent protection device is connected in series between the main power input terminal and the first terminal of the common pre-charge resistor. The overcurrent protection device is an integrated electronic fuse chip. The fault status output terminal of the overcurrent protection device is connected to the control unit. N precharge branches, where N is an integer greater than 1. Each of the N precharge branches includes a precharge selection switch, which is a semiconductor switch. The first terminal of each precharge selection switch is connected to the second terminal of a common precharge resistor, and the second terminal of each precharge selection switch is connected to the power supply terminal of the corresponding load. N main power supply branches, each of the N main power supply branches includes a main power supply switch. The first end of the main power supply switch is connected to the main power input terminal, and the second end of the main power supply switch is connected to the power supply terminal of the corresponding load. N-channel voltage detection circuits, where the detection terminal of each voltage detection circuit is connected to the power supply terminal of the corresponding load to detect the voltage on the load side. as well as The control unit is electrically connected to the control terminals of each precharge selection switch, each main power supply switch, and the output terminal of each voltage detection circuit. The control unit is configured to: sequentially select each precharge selection switch in a preset order, so that the common precharge resistor limits the current charging of each load-side capacitor at different time periods, and at most only one precharge selection switch is in the on state at any given time; when a voltage detection circuit of the N voltage detection circuits detects that the voltage on the load side of that circuit reaches a preset threshold, close the main power supply switch of that circuit and open the precharge selection switch of that circuit, so that the common precharge resistor is disconnected from the precharge branch of that circuit. When the continuous conduction time of a certain precharge selection switch exceeds the preset maximum precharge time and the load side voltage of that circuit does not reach the preset threshold, the precharge selection switch of that circuit is disconnected, the precharge of that circuit is stopped, and the main power supply switch of that circuit is not closed.

2. The power supply system with multiple alternating pre-charging as described in claim 1, characterized in that, The preset threshold is a preset ratio of the input voltage at the main power input terminal. The control unit is also configured to determine that the pre-charging of a certain load side is complete and close the main power supply switch of that line only when the voltage of a certain load side reaches the preset threshold for multiple consecutive sampling cycles.

3. The power supply system with multiple alternating pre-charging as described in claim 1, characterized in that, The overcurrent protection device is configured to have two levels of overcurrent protection: the first level is short-circuit protection, with an overcurrent threshold much higher than the total current of each load when it is operating normally, and a response time in the microsecond range; the second level is continuous overcurrent protection, with an overcurrent threshold higher than the total current of each load when it is operating normally but lower than the overcurrent threshold of the first level, and a response time in the millisecond range.

4. The power supply system with multiple alternating pre-charging as described in claim 1, characterized in that, The control unit is also configured to support dynamic configuration of the selection order of each of the precharge selection switches, so as to give priority to precharging of preset key load branches.

5. The multi-channel alternating pre-charging power supply system according to claim 4, characterized in that, The control unit is also configured to: when the system is cold-started and no external pre-charge sequence configuration instruction is received, select each of the pre-charge selection switches in sequence according to the locally preset pre-charge sequence. The locally preset pre-charge sequence is: first pre-charge the load branch where the control module and sensor are located, and then pre-charge the load branch where each joint motor is located in sequence.

6. The power supply system with multiple alternating pre-charging as described in claim 1, characterized in that, The power supply system is for a humanoid robot. The main power input terminal receives a 48V DC voltage, N equals 5, and the five loads correspond to the power supply paths of the joint motors of the five body parts of the humanoid robot.

7. The power supply system with multiple alternating pre-charging as described in claim 1, characterized in that, Each of the precharge selection switches and the corresponding voltage detection circuits are integrated into an independent precharge function module. The power supply system is configured such that when the number of power supply circuits is increased, only the precharge function module needs to be added and connected to the second terminal of the common precharge resistor and the control unit, without replacing the common precharge resistor.

8. A multi-channel alternating pre-charging method, characterized in that, A power supply system for multiple alternating pre-charging as described in any one of claims 1 to 7, the power supply system comprising a main power input terminal, a common pre-charging resistor, N pre-charging branches, N main power supply branches, N voltage detection circuits, and a control unit, wherein N is an integer greater than 1; the first terminal of the common pre-charging resistor is connected to the main power input terminal; each pre-charging branch includes a pre-charging selection switch, the pre-charging selection switch being a semiconductor switch; the first terminals of all pre-charging selection switches are commonly connected to the second terminal of the common pre-charging resistor; the second terminals of each pre-charging selection switch are respectively connected to the power supply terminal of a corresponding load; each main power supply branch includes a main power supply switch; the first terminal of the main power supply switch is connected to the main power input terminal; the second terminal of the main power supply switch is connected to the power supply terminal of a corresponding load; the detection terminal of each voltage detection circuit is connected to the power supply terminal of a corresponding load; the control unit is electrically connected to the control terminal of each pre-charging selection switch, the control terminal of each main power supply switch, and the output terminal of each voltage detection circuit; the method includes: Step S1: System initialization, control all the precharge selection switches and all the main power supply switches to be in the off state; Step S2: Select the first precharge selection switch in a preset order so that the common precharge resistor limits the current of the load-side capacitor of that channel, and at most only one precharge selection switch is in the on state at any given time. Step S3: The voltage detection circuit of this circuit detects the load side voltage in real time. When the load side voltage reaches the preset threshold, the main power supply switch of this circuit is closed and the precharge selection switch of this circuit is opened to release the common precharge resistor. Step S4: Select the next precharge selection switch according to the preset order, and repeat step S3 until all N loads have completed precharging.

Images