A physical execution control chip and method based on the energy evolution of seven mechanical components
By using a simulated circuit structure based on the energy evolution of seven mechanical components, the problems of digital delay, storage-computation separation, and security in existing physical execution control chips are solved, achieving efficient and safe physical work control, which is suitable for high dynamic systems such as robots.
Patent Information
- Application Number
- CN202611293038.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-25
- Publication Date
- 2026-11-17
AI Technical Summary
Existing physical execution control chips suffer from problems such as digital latency, separation of storage and computation, reliance on software for security mechanisms, and inability to achieve dynamic equilibrium of physical scale.
It adopts a simulated circuit structure with seven mechanical components for energy evolution, executes the control algorithm directly at the physical level, uses components such as RC network, analog multiplier, and memristor array to realize the output of physical work signal, and provides intrinsic safety protection through hard limiting circuit.
It achieves nanosecond-level physical convergence, inherent physical safety, and fully parallel multi-path control, making it suitable for synchronous control of highly dynamic systems such as robots, and possessing cross-scenario universality.
Smart Images

Figure CN123113337A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the interdisciplinary field of integrated circuit design, analog computing, and automatic control, specifically relating to a physical execution control chip and method based on the energy evolution of seven mechanical components. This invention can serve as a physical control-level hardware coprocessor, widely applicable to embodied intelligence and industrial control scenarios requiring low latency, physical non-bypass capabilities, and high dynamic response speeds, such as robot joint actuation, autonomous vehicle stability control, and industrial process temperature control. Background Technology
[0002] Existing physical execution control technologies typically employ a digital control architecture consisting of a sensor, ADC, microcontroller (MCU / CPU), DAC, and power amplifier. This architecture suffers from the following fundamental drawbacks: (1) Computational delay and discreteness: Digital control requires steps such as sampling, quantization, and software instruction execution, which are limited by clock frequency. For highly dynamic physical systems that require nanosecond-level response (such as robot fall recovery), the delay of digital control cannot be ignored.
[0003] (2) The “memory wall” bottleneck of memory-computation separation: control parameters and state variables are frequently moved between memory and processor, resulting in increased power consumption and difficulty in achieving multi-channel synchronous parallel control.
[0004] (3) Security mechanisms rely on software: Traditional overheat / overload protection relies on software code judgment, which may be tampered with or bypassed by attackers (such as "jailbreaking"), and cannot achieve absolute security at the physical level.
[0005] (4) Lack of "dynamic equilibrium" control based on physical dynamics: Since the control algorithm runs in the discrete time domain, traditional PID or MPC is prone to overshoot or oscillation. The system lacks a natural convergence mechanism based on the physical dissipative structure characteristics and cannot achieve dynamic equilibrium control that automatically regresses to the "neutral steady state (0.5)".
[0006] The applicant's earlier application (CN202610601939.6) discloses a cognitive chip and cognitive data processing method based on event relation grammar. However, this chip is essentially "virtual and real work done at the cognitive layer" and does not have the ability to directly drive physical actuators to do work. The applicant's earlier application, "A System State Control Method and System Based on the Energy Evolution of Seven Mechanical Components," proposes a general control architecture at the software algorithm level, but does not provide a complete technical solution for directly imprinting the seven mechanical component equations into the analog hardware circuit and converting them into "physical work signals."
[0007] Therefore, there is an urgent need for a physical execution control chip that can directly execute the seven mechanical component equations at the physical analog circuit level, directly convert physical state deviations into driver power signals, and has an inherent physical safety mechanism. Summary of the Invention
[0008] (a) Purpose of the invention The purpose of this invention is to provide a physical execution control chip and method based on the energy evolution of seven mechanical components, so as to solve the technical problems of existing physical control chips, such as digital latency, separation of storage and computation, reliance on software for security mechanisms, and inability to achieve dynamic equilibrium of physical scale.
[0009] Specific objectives include: (1) A chip is provided that implements the seven mechanical component equations in a pure analog circuit and directly outputs the physical work signal.
[0010] (2) Provide a normalized physical interface that directly connects the physical sensor input voltage to a 0.5V neutral reference voltage.
[0011] (3) A circuit structure is provided that uses RC network, analog multiplier, memristor array and other components to physically implement the seven mechanical components respectively.
[0012] (4) Provides an intrinsic security regulation mechanism that requires no clock, no software code, and relies entirely on the physical characteristics of the circuit.
[0013] (II) Technical Solution To achieve the above objectives, the present invention provides the following technical solution: A physical execution control chip based on the energy evolution of seven mechanical components, characterized in that it includes: Physical state input interface: used to receive analog voltage signals from sensors (such as IMU, temperature sensor) of the controlled physical system; the chip has a built-in normalization circuit to accurately map the "physical zero deviation state" to a normalized reference voltage of 0.5V.
[0014] Neutral reference voltage source: Used to provide a high-precision, low-temperature-drift 0.5V reference voltage.
[0015] A seven-component mechanical component simulation evolution circuit array is used to execute the energy evolution equations of the seven mechanical components in real time in the simulation domain. The circuit array is based on Kirchhoff's current law and the characteristics of the analog multiplier, and drives the input voltage to converge to a 0.5V neutral reference voltage through the natural response of the physical circuit. The seven mechanical components include coupling term, taiji term, sharpening term, awareness term, source term, saturation term, and disturbance term.
[0016] Physical work output interface: includes a transconductance amplifier (OTA) to proportionally amplify the converged 0.5V reference voltage difference (physical deviation) in the evolution circuit into a high current output, which directly drives the actuator (motor, heater, etc.) to achieve physical work.
[0017] Furthermore, the specific circuit implementation of the seven mechanical component simulation evolution circuit array is as follows: - Taiji term and source term (regression and damping): Implemented using an RC discharge network circuit, the capacitor discharge time constant is determined by parameters (β and ζ), and the driving voltage regresses to 0.5V; - Sharpening term (gain amplification): Implemented using a square root / cubic root nonlinear amplifier composed of analog multipliers; - Saturation term (nonlinear limiting): Three analog multipliers are cascaded to perform cubic calculations, which are then converted into current injections through a transimpedance amplifier to prevent high-frequency oscillations near 0.5V. - Awareness (Collaboration): The average value (ē) of each voltage is calculated using a resistor network and operational amplifier, and then a feedback signal is generated through a transconductance amplifier; - Coupling terms (gravity-dominated): A 6×6 coupling matrix is constructed using a programmable memristor cross array; the programmable memristors adjust the coupling relationships between each state dimension in real time; - Disturbance term (external sensing): Random voltage disturbances are provided using an on-chip thermal noise generator to simulate external physical shocks.
[0018] Furthermore, the chip also includes a hard limit protection circuit: it is directly connected to the evolution circuit array. When the evolution voltage deviates from the 0.5V reference by more than a preset threshold (exceeding the safety range), it directly forces the output current to be pulled down to zero through an analog switch, thereby achieving physical-level unbypassable safety protection.
[0019] (III) Beneficial Effects (1) Real physical work: The present invention outputs a high current execution signal, which directly drives the actuator, realizing "physical output" rather than the digital encoding output of existing cognitive chips.
[0020] (2) Nanosecond-level physical convergence: The entire evolution process relies entirely on the physical response of the analog circuit (charge charging and discharging time), and there is no software instruction cycle, realizing true "electricity-controlled electricity".
[0021] (3) Intrinsic physical security: The hard limiting circuit directly blocks the physical work current at the circuit level. Its response speed is determined by the physical delay of the analog switch, which is extremely fast and absolutely cannot be tampered with by software.
[0022] (4) Decentralized control: No clock, no CPU instructions, multi-path physical control can evolve in parallel, naturally suitable for the synchronous control of multiple joints of a robot.
[0023] (5) Engineering universality: The control architecture proposed in this invention deeply follows the natural law of physical systems converging towards a steady state. Its internal mechanism is highly consistent with the concepts of interaction (coupling term), robustness (source term / Taiji term), disturbance resistance (disturbance term), and stability (saturation term) of various parts of the system in classical control theory. Therefore, it has strong cross-scenario universality. Attached Figure Description
[0024] Figure 1 : Overall architecture diagram of the physical execution control chip described in this invention.
[0025] Figure 2 : A schematic diagram of the macroscopic structure of a circuit array that simulates the evolution of seven mechanical components.
[0026] Figure 3 : Detailed circuit diagram of Taiji term and source term (RC discharge network).
[0027] Figure 4 : Detailed circuit schematics of the sharpening and saturation terms (analog multiplier array).
[0028] Figure 5 Connection diagram of the physical work output interface (transconductance amplifier) and the actuator.
[0029] Figure 6 : Structure diagram of hard limiting protection circuit. Detailed Implementation
[0030] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention.
[0031] Principle verification explanation (no data fabrication statement): The core algorithm of the chip described in this invention—the seven-component energy evolution equation—has been fully calculated and verified in the applicant's prior research result, the WOLM cognitive decision engine (the core algorithm of which is open-source on GitHub, repository address: https: / / github.com / WOLM9123 / wolm). This verification system fully implements the same seven-component core algorithm as this invention at the software level. It should be specifically noted that the WOLM software verifies the mathematical logic correctness of the seven-component algorithm, not the physical measurement data at the hardware level of this chip. Since this chip has not yet been fabricated, all descriptions of hardware performance in this application are based on circuit theory calculations (such as RC time constant, operational amplifier bandwidth, etc.), and no untested or fabricated chip measurement data are provided. Those skilled in the art can reproduce the technical solution of this invention based on the following circuit principles.
[0032] Example 1: Basic Circuit Structure of a Chip The chip's physical state input interface includes six input terminals for receiving six-dimensional physical quantities. The normalization circuit, through proportional amplification and offset adjustment, precisely maps a physical state with "zero deviation" to 0.5V. The neutral reference voltage source employs a bandgap reference circuit, outputting a high-precision 0.5V.
[0033] The core of the seven-component mechanical simulation evolution circuit array is composed of general-purpose analog electronic components: S1. For the taiji term (β·(0.5-e)) and the source term (ζ·(0.5-e)): In circuit theory, these two terms can be implemented using a precision resistor network connected from the node to a 0.5V reference voltage and then to ground. According to Kirchhoff's current law, the charging current generated by this network to the 0.5V reference is proportional to (0.5-e), and its time constant is determined by the resistance and capacitance values.
[0034] S2. For the sharpening term: exponential nonlinear amplification can be achieved through an analog multiplier in conjunction with a signal processing circuit. When the voltage deviates significantly from 0.5V, the sharpening circuit generates a large current to accelerate convergence to the reference; when the deviation is small, the output is almost zero.
[0035] S3. For the saturation term (-κ·(e-0.5)³): The cubic operation is performed by cascading three analog multipliers, and the result is fed back to the node through an inverting amplifier. Theoretically, this circuit generates almost no feedback current when the voltage is close to 0.5V (with minimal deviation), thus greatly reducing the occurrence of high-frequency oscillations.
[0036] S4. For the coupling term (α·J·(e-0.5)): A programmable memristor cross array is used to implement 6×6 matrix multiplication, and the coupling weights between physical quantities can be adjusted according to external configuration (written through the configuration interface). When the robot is supported by one leg or two legs, the coupling coefficient can be increased or decreased in real time.
[0037] Example 2: Digital Calibration Configuration Steps When the chip is used for the first time or the actuator is replaced, a digital calibration configuration step needs to be performed. This specifically includes: Preset circuit parameters (including coupling coefficient α, taiji coefficient β, sharpening coefficient γ, perception coefficient η, source coefficient ζ, saturation coefficient κ, and disturbance intensity d) are written into the on-chip programmable resistor array via an external interface or on-chip configuration unit. The on-chip programmable resistor array adjusts the corresponding resistance value according to the written digital configuration value to match the characteristics of the current physical system (such as motors with different inertia or heaters with different heat capacities).
[0038] In this step, if the physical execution control system is connected to different types of actuators, the digital calibration configuration parameters are calculated in an off-chip or on-chip processor using a digital calibration algorithm and written through the external interface. The digital calibration configuration step, in conjunction with the normalization circuit of the physical state input interface, ensures that the physical zero-deviation state is accurately mapped to a 0.5V reference voltage.
[0039] Example 3: Physical work process (taking robot balance control as an example) When the robot is in a state of physical inclination to tip over, the deviation voltage generated by the torque sensors at the joints is input to the chip through the physical state input interface. The normalization circuit maps the "current tilt angle deviation" to a voltage value, assuming an offset of 0.5V from the reference (e.g., 0.3V).
[0040] The voltage enters the seven-component mechanical simulation evolution circuit array, and the circuit begins its physical evolution. The disturbance term senses external wind disturbances; the coupling term performs collaborative calculations based on the bipedal support state; the sharpening and saturation terms work together, and after an extremely short physical time constant (calculated based on RC parameters), the voltage automatically converges back to 0.5V.
[0041] The convergence voltage of the 0.5V reference is input to the transconductance amplifier. The transconductance amplifier amplifies the "current deviation voltage" into "drive current," which is directly output to the motor driver. The motor generates a reverse torque, the robot's tilt angle decreases until the sensor deviation reaches zero, the chip voltage stabilizes at 0.5V, the motor stops, and the robot achieves dynamic equilibrium.
[0042] Example 4: Physical Intrinsic Safety Protection The chip's internal hard-limiting protection circuit includes a high-speed voltage comparator. Its first input is connected to the evolution circuit node, and its second input is connected to a preset safety threshold voltage. If the evolution node voltage briefly deviates from the safety threshold by 0.5V due to a severe surge (e.g., exceeding 1.0V or falling below 0.0V), the comparator output flips, directly disconnecting the physical power output interface switch and instantly cutting off the motor current. The entire process involves no software execution and no instruction delay.
[0043] The above description is merely a preferred embodiment of the present invention and does not limit the scope of protection of the present invention. Any equivalent structural or procedural modifications made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of protection of the present invention.
Claims
1. A physical execution control chip based on the energy evolution of seven mechanical components, characterized in that, include: The physical state input interface is used to receive the physical state analog voltage signal from the controlled physical system and map the physical zero deviation state to a 0.5V reference voltage. A neutral reference voltage source is used to provide a high-precision 0.5V reference voltage; A seven-component mechanical component simulation evolution circuit array is used to execute the energy evolution equation of seven mechanical components, including coupling term, taiji term, sharpening term, awareness term, source term, saturation term and perturbation term, in real time in the analog domain based on Kirchhoff's current law and the characteristics of the analog multiplier, driving the physical state analog voltage signal to naturally converge to the 0.5V reference voltage; The physical work output interface includes a transconductance amplifier, which is used to proportionally amplify the converged deviation voltage into a large current signal to directly drive the physical actuator to do work.
2. The chip according to claim 1, characterized in that, The seven-component mechanical component simulation evolution circuit array includes: An RC discharge network circuit is used to implement the Taiji term and the source term; An analog multiplier nonlinear amplifier is used to implement the sharpening term; Three cascaded analog multipliers and transimpedance amplifiers are used to implement the saturation term; A resistor network and an operational amplifier are used to calculate the mean of each dimension to achieve the perceived item; A programmable memristor cross array is used to implement the coupling term; An on-chip thermal noise generator is used to implement the disturbance term.
3. The chip according to claim 1, characterized in that, The physical work output interface includes a transconductance amplifier (OTA), whose output current is proportional to the input voltage; the high current output terminal of the OTA is connected to the drive terminal of the actuator.
4. The chip according to claim 1, characterized in that, It also includes a hard limit protection circuit, which is directly connected between the seven mechanical component simulation evolution circuit array and the physical work output interface; when the evolution voltage deviates from 0.5V and exceeds the preset safety threshold, the output current is forcibly pulled down to zero through the analog switch.
5. The chip according to claim 1, characterized in that, The programmable memristor cross array is a 6×6 array, and its resistance value can be programmed online through the configuration interface to dynamically adjust the coupling weight between different physical dimensions; when the controlled system is in a high coupling mode, the resistance value is programmed to increase the coupling strength.
6. A physical execution control method based on the chip according to any one of claims 1 to 5, characterized in that, Includes the following steps: Step S1: Receive the state analog voltage signal of the physical system through the physical state input interface and normalize it to a 0.5V reference. Step S2: The seven mechanical component simulation evolution circuit array receives the voltage signal and executes the seven mechanical component equations in the analog domain through RC charging and discharging, analog multiplier nonlinear operation and memristor coupling, so that the voltage signal naturally evolves and converges to the 0.5V reference. Step S3: After the voltage converges, the transconductance amplifier of the physical work output interface converts it into a large current signal to drive the physical actuator to complete the physical work action. Step S4: The hard limit protection circuit monitors the evolving voltage in real time. If it deviates from the threshold, it instantly cuts off the current output to achieve intrinsic physical safety.
7. The method according to claim 6, characterized in that, In step S1, the specific implementation of mapping the physical zero deviation state to a 0.5V reference is as follows: through a proportional amplification and offset adjustment circuit, the physical deviation voltage signal output by the sensor is converted into a normalized voltage signal centered at 0.5V.
8. The method according to claim 6, characterized in that, In step S2, the seven mechanical component equations are specifically: Δe = α·J·(e-0.5) + β·(0.5-e) + γ·sign(e-0.5)·|e-0.5|^1.5 + η·(e-ē) + ζ·(0.5-e) - κ·(e-0.5)³ + d·ξ, where e is the physical state voltage vector, J is a 6×6 coupling matrix, ē is the mean of e, ξ is a six-dimensional random perturbation vector, α, β, γ, η, ζ, and κ are preset circuit parameters, and d is the perturbation intensity; the circuit parameters can be configured through an on-chip programmable resistor array.
9. A physical execution control system based on the energy evolution of seven mechanical components, characterized in that, include: The physical execution control chip as described in any one of claims 1 to 5; And at least one physical actuator, the physical actuator being connected to the physical work output interface of the physical execution control chip; The physical execution control chip directly drives the actuator to operate based on the state feedback of the controlled physical system.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of acquiring and writing digital calibration configuration parameters in the method as described in any one of claims 6 to 8.
Citation Information
Patent Citations
Cognitive chip based on event relation grammar and cognitive data processing method
CN122197786A
