An intelligent cockpit unified safety control method and device, a vehicle, and a storage medium

Abstract

The application discloses a kind of intelligent cockpit unified safety control method, device, vehicle and storage medium.It relates to the technical field of intelligent cockpit, method includes: obtaining the multi-source perception data in intelligent cockpit, time alignment is carried out to the multi-source perception data, and time sequence input sequence set is obtained;The time sequence input sequence set includes current time sequence input data and the historical time sequence input data corresponding to current time sequence input data;From the time sequence input sequence set, select multiple time sequence input data and input to time sequence encoder, obtain current state embedding vector;Determine current continuous risk value based on the pre-constructed risk mapping function and the current state embedding vector;Based on the current continuous risk value, each module request in current module request set is sorted, and execution strategy set is obtained;Based on the execution strategy set, corresponding control instruction is generated for each functional module.The method can uniformly manage each functional module in intelligent cockpit.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of smart cockpit technology, and in particular to a unified safety control method, device, vehicle and storage medium for smart cockpits. Background Technology

[0002] As the integration of intelligent cockpit functions continues to improve, in-vehicle interaction systems involve multiple subsystems, including infotainment, navigation, communication, and driver assistance. The concurrent flow of information from multiple sources increases the cognitive load on the driver, and fluctuations in the driver's attention can directly impact driving safety.

[0003] In existing technologies, cockpit interaction control typically adopts a method of maintaining the priority of each module independently, or makes classification decisions based on discrete risk levels. It lacks a unified global risk measurement space and cross-module coordination mechanism, making it difficult to adapt to dynamically changing driving environments. Summary of the Invention

[0004] This invention provides a unified safety control method, device, vehicle, and storage medium for intelligent cockpits to solve the problem that existing cockpit control methods cannot adapt to dynamically changing driving environments.

[0005] According to one aspect of the present invention, a unified safety control method for a smart cockpit is provided, the method comprising: Acquire multi-source sensing data from the intelligent cockpit, perform time alignment on the multi-source sensing data, and obtain a time-series input sequence set; the time-series input sequence set includes the current time-series input data and the historical time-series input data corresponding to the current time-series input data. Multiple time-series input data are selected from the time-series input sequence set and input into the time-series encoder to obtain the current state embedding vector; The current continuous risk value is determined based on the pre-constructed risk mapping function and the current state embedding vector; Based on the current continuous risk value, the requests of each module in the current module request set are sorted to obtain the execution strategy set; Based on the set of execution strategies, corresponding control instructions are generated for each functional module.

[0006] According to another aspect of the present invention, a unified safety control device for intelligent cockpits is provided, the device comprising: The acquisition module is used to acquire multi-source sensing data in the intelligent cockpit, perform time alignment on the multi-source sensing data, and obtain a time-series input sequence set; the time-series input sequence set includes the current time-series input data and the historical time-series input data corresponding to the current time-series input data. The input module is used to select multiple time-series input data from the time-series input sequence set and input them into the time-series encoder to obtain the current state embedding vector; The determination module is used to determine the current continuous risk value based on a pre-constructed risk mapping function and the current state embedding vector; The sorting module is used to sort the module requests in the current module request set based on the current continuous risk value, so as to obtain an execution strategy set; The generation module is used to generate corresponding control instructions for each functional module based on the set of execution strategies.

[0007] According to another aspect of the present invention, a vehicle is provided, the vehicle comprising: at least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the unified safety control method for smart cockpits according to any embodiment of the present invention.

[0008] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the unified safety control method for intelligent cockpits according to any embodiment of the present invention.

[0009] This invention discloses a unified safety control method, device, vehicle, and storage medium for an intelligent cockpit. The method includes: acquiring multi-source sensing data from the intelligent cockpit; performing time alignment on the multi-source sensing data to obtain a time-series input sequence set; the time-series input sequence set includes current time-series input data and historical time-series input data corresponding to the current time-series input data; selecting multiple time-series input data from the time-series input sequence set and inputting them into a time-series encoder to obtain a current state embedding vector; determining a current continuous risk value based on a pre-constructed risk mapping function and the current state embedding vector; sorting the module requests in a current module request set based on the current continuous risk value to obtain an execution strategy set; and generating corresponding control instructions for each functional module based on the execution strategy set. This method obtains the current continuous risk value by analyzing multi-source sensing data in the intelligent cockpit and determines how to control each functional module based on the risk value, thereby enabling unified control of the intelligent cockpit and solving the problem that existing cockpit control methods cannot adapt to dynamically changing driving environments.

[0010] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0012] Figure 1 This is a flowchart illustrating a unified safety control method for an intelligent cockpit provided in Embodiment 1 of the present invention. Figure 2 A schematic diagram of a timing encoder provided in an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the priority ordering of module requests, provided as an embodiment of the present invention. Figure 4 A schematic diagram of risk status adjustment provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of the structure of a unified safety control device for an intelligent cockpit provided in Embodiment 2 of the present invention; Figure 6 This is a structural diagram of a vehicle according to the unified safety control method for intelligent cockpits in an embodiment of the present invention. Detailed Implementation

[0013] To enable those skilled in the art to better understand the present invention, 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 merely some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention. It should be understood that the various steps described in the method embodiments of the present invention can be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.

[0014] The term "comprising" and its variations as used herein are open-ended inclusions, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Definitions of other terms will be given in the description below.

[0015] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, any variations of the terms "comprising" and "having," etc., are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0016] It should be noted that the terms "a" and "a plurality of" used in this invention are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0017] The names of the messages or information exchanged between the multiple devices in the embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of these messages or information.

[0018] It is understood that before using the technical solutions disclosed in the various embodiments of this disclosure, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this disclosure in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained.

[0019] For example, upon receiving a user's active request, a prompt message is sent to the user to explicitly inform them that the requested operation will require the acquisition and use of the user's personal information. This allows the user to independently choose whether to provide personal information to the software or hardware, such as the electronic device, application, server, or storage medium performing the operations of this disclosed technical solution, based on the prompt message.

[0020] As an optional but non-limiting implementation, in response to a user's active request, sending a prompt message to the user can be done via a pop-up window, where the prompt message can be presented in text format. Furthermore, the pop-up window can also include a selection control allowing the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0021] It is understood that the above notification and user authorization process are merely illustrative and do not constitute a limitation on the implementation of this disclosure. Other methods that comply with relevant laws and regulations may also be applied to the implementation of this disclosure.

[0022] It is understood that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and related provisions.

[0023] With the iteration of artificial intelligence (AI) algorithms and the upgrading of cockpit chip capabilities, vehicle cockpits can integrate multimodal data and AI models, and then form effective control outputs through unified processing by the cockpit system. Therefore, this invention proposes a unified safety control method for intelligent cockpits based on continuous risk field modeling.

[0024] Example 1 Figure 1 This is a flowchart illustrating a unified safety control method for an intelligent cockpit according to Embodiment 1 of the present invention. This method is applicable to situations where various modules in an intelligent cockpit are controlled in a unified manner. This method can be executed by a unified safety control device for an intelligent cockpit, which can be implemented by software and / or hardware and is generally integrated into a vehicle. In this embodiment, the vehicle includes, but is not limited to, large vehicles, small vehicles, and medium-sized vehicles.

[0025] like Figure 1 As shown in Embodiment 1 of the present invention, a unified safety control method for an intelligent cockpit includes the following steps: S110. Acquire multi-source sensing data in the intelligent cockpit, perform time alignment on the multi-source sensing data, and obtain a time-series input sequence set; the time-series input sequence set includes the current time-series input data and the historical time-series input data corresponding to the current time-series input data.

[0026] Here, "intelligent cockpit" refers to the cockpit of an autonomous vehicle. "Multi-source sensing data" refers to sensing data from different sources. "Temporal input sequence set" can be a collection containing multiple temporal input data, which can refer to the feature set of multi-source input data at a certain moment. The temporal input sequence set can include the current temporal input data and the corresponding historical temporal input data. The current temporal input data can refer to the temporal input data corresponding to the current moment, and the historical temporal input data can refer to the temporal input data before the current moment.

[0027] In this embodiment, multi-source sensing data from the smart cockpit can be acquired, and the multi-source sensing data can be time-aligned to obtain a time-series input sequence set.

[0028] In one embodiment, acquiring multi-source perception data in the smart cockpit includes: collecting data in the smart cockpit at a uniform sampling frequency through an onboard sensor network to obtain multi-source perception data; the multi-source perception data includes at least driver attention feature data, driver interaction behavior data, and vehicle motion dynamic data.

[0029] Among them, the vehicle sensor network can be a distributed sensing and acquisition network built inside the vehicle, which can collect data from the driver and the vehicle.

[0030] In this embodiment, the multi-source perception data includes at least driver attention feature data, driver interaction behavior data, and vehicle motion dynamic data. These three types of data can be collected at a uniform sampling frequency through an onboard sensor network. By uniformly representing the driver's attention state and the vehicle's operating state and mapping it to continuous risk values, unified measurement and global control across functional modules can be achieved.

[0031] Driver attention characteristic data refers to data related to a driver's attention while driving a vehicle. For example, driver attention characteristic data may include eye movement vectors, fixation time, pupil diameter, head posture, and eye opening / closing angle acquired by a Driver Monitoring System (DMS). Driver interaction behavior data refers to data generated when a driver interacts with in-vehicle devices. For example, driver interaction behavior data may include the location coordinates and operation types of touch events generated by interactions with the central control screen, steering wheel buttons, etc. Vehicle dynamic data refers to vehicle operation data. For example, vehicle motion dynamic data may include vehicle speed, longitudinal and lateral acceleration, steering wheel angle and rate of turn, turn signal status, brake pressure, and advanced driver assistance system warning signals acquired via the controller area network bus.

[0032] All the above data are timestamped and preprocessed to form a time-series input sequence set S.

[0033] X_ t =[F att ,F int ,F dym ] t ; S={X (t-n△t) ,...,X_ t}; Where t represents the current time, X _t As a vector, it represents the set of multi-source input data at time t, where Δt is the sampling time interval, which can be 33ms based on 30 frames from the DMS camera; F att F represents driver attention characteristic data. int This represents driver interaction data, including touch coordinates, button operation frequency, voice interaction status, and operation type, etc. dym S represents the dynamic data of vehicle operation, and S represents the sequence set of X_t in the time dimension.

[0034] S120. Select multiple timing input data from the timing input sequence set and input them into the timing encoder to obtain the current state embedding vector.

[0035] The temporal encoder can be a pre-trained encoder; this embodiment does not limit the structure of the temporal encoder. The current state embedding vector can refer to the embedding vector obtained based on the data encoding of the current vehicle.

[0036] In this embodiment, the timing input sequence set may include multiple timing input sequences. Multiple timing input data can be selected from the timing input sequence set and input to the timing encoder to obtain the current state embedding vector.

[0037] In one embodiment, the step of selecting multiple time-series input data from the time-series input sequence set and inputting them into a time-series encoder to obtain a current state embedding vector includes: defining a time window length; selecting current time-series input data from the time-series input sequence set based on the time window length, and historical time-series input data within the time window as the current input sequence; and inputting the current input sequence into the time-series encoder to obtain the current state embedding vector.

[0038] The time window length can be the length of the time window for selecting the timing input data, and the time window length can be determined based on the timing encoder. The current input sequence can be the timing input sequence input to the timing encoder at the current time.

[0039] In this embodiment, the time window length can be predefined. The time window starts from the current time-series input data. Based on the time window length, the current time-series input data and all historical time-series input data within the time window are selected from the time-series input sequence set as the current input sequence. The current input sequence is then input to the time encoder to obtain the current state embedding vector.

[0040] For example, Figure 2 A schematic diagram of a timing encoder provided for an embodiment of the present invention is shown below. Figure 2 As shown, the time encoder has at least hidden layer 1 and hidden layer 2. The time window length k is defined, and the input sequence X_{tk:t} is formed by taking the current time t and the time-series input data of the k1 previous times. This input sequence is input into the time encoder, and the generated high-dimensional feature vector is used as the current state embedding vector S_t, expressed as: S_t=Encoder(X_{tk:t}); Where S_t represents the state embedding vector, which is a numerical representation in a high-dimensional continuous space, implicitly containing the temporal coupling relationship between the driver's state (driver attention feature data, driver interaction behavior data) and the vehicle's operating state (vehicle motion dynamic data); Encoder(·) represents the temporal encoding function, which can be implemented using a long short-term memory network or an attention-based transformer model; X_{tk:t} represents the input data sequence within a time window of length k; k is the length of the time window, corresponding to a preset time span.

[0041] S130. Determine the current continuous risk value based on the pre-constructed risk mapping function and the current state embedding vector.

[0042] The risk mapping function can be a pre-built function that maps the relationship between the current cabin state and the risk value. The current continuous risk value can be the continuous risk value at the current moment, and the continuous risk value is a real number.

[0043] In this embodiment, the current continuous risk value can be determined based on a pre-constructed risk mapping function and the current state embedding vector. Mapping the driver's attention state and the vehicle's operating state to continuous risk values ​​makes the interaction behaviors of different functional modules comparable and calculable under the same risk scale, fundamentally solving the problem of fragmented control mechanisms.

[0044] In one embodiment, determining the current continuous risk value based on a pre-constructed risk mapping function and the current state embedding vector includes: determining the rate of change of the current state embedding vector based on the current state embedding vector and the state embedding vector at the previous moment; and inputting the current state embedding vector and the rate of change into the risk mapping function to obtain the current continuous risk value.

[0045] In this embodiment, the rate of change of the current state embedding vector can be calculated based on the current state embedding vector and the state embedding vector at the previous moment. By inputting the current state embedding vector and the rate of change into the risk mapping function, the current continuous risk value can be obtained.

[0046] For example, a risk mapping function f(·) is defined, which, combined with the current state embedding vector S_t and its rate of change ΔS_t, calculates the continuous risk value R_t, expressed as: R_t=f(S_t,ΔS_t) ΔS_t=S_t-S_{t1} Where R_t is the continuous risk value at the current moment, taking the value of a real number, forming a unified risk measurement space; f(·) is the risk mapping function, used to map the state embedding vector and its rate of change to the risk value; ΔS_t is the time rate of change of the state embedding vector, reflecting the evolution trend of the state; S_{t1} is the state embedding vector at the previous moment. By replacing discrete risk levels with continuous risk values, the system behavior can be finely controlled.

[0047] In this embodiment, the risk mapping function f(·) can also introduce a user-personalized parameter θ_user to form a parameterized risk function, expressed as: R_t=f(S_t,ΔS_t,θ_user); Here, θ_user is a personalized parameter learned based on the driver's historical interaction behavior data, used to adapt to the attention characteristics and risk preferences of different drivers.

[0048] S140. Based on the current continuous risk value, sort the module requests in the current module request set to obtain the execution strategy set.

[0049] The current module request set refers to the collection of module requests from all functional modules within the current cockpit. A functional module can be a module that performs a specific function within the cockpit, such as a music playback module, a navigation module, or an autopilot module. A module request can be an interaction request initiated by a functional module. The execution strategy set can be a set of module requests arranged in a specific order, and the system can process module requests according to the order of the execution strategy set.

[0050] In this embodiment, the requests from each module in the current module request set can be sorted based on the current continuous risk value, with important module requests placed at the front and less important ones at the back, resulting in an execution strategy set. System regulation based on continuous risk values ​​can avoid abrupt control changes and user experience disruptions caused by discrete level switching, making the system interaction process smoother.

[0051] For example, Figure 3 This is a schematic diagram illustrating the priority ordering of module requests provided in an embodiment of the present invention, as shown below. Figure 3 As shown, Advanced Driving Assistance System (ADAS) requests have a higher priority than music and navigation requests.

[0052] In one embodiment, the step of sorting the module requests in the current module request set based on the current continuous risk value to obtain an execution strategy set includes: obtaining the current module request set in the intelligent cockpit at the current moment; the module request set includes module requests of multiple functional modules and the original priority corresponding to each module request; for each functional module, adjusting the original priority of the module request based on the current continuous risk value and the risk sensitivity function corresponding to the functional module to obtain the adjusted priority; and sorting the module requests of each functional module based on the adjusted priority to obtain the current execution strategy set.

[0053] The current module request set can include module requests from multiple functional modules and the original priority of each module request. The original priority can be a pre-set priority. The risk sensitivity function can be a pre-set function relating risk value and priority. Different functional modules can have different risk sensitivity functions.

[0054] In this embodiment, the current set of module requests within the intelligent cockpit at the current moment can be obtained. For each functional module, based on the current continuous risk value and combined with the risk sensitivity function corresponding to that functional module, the original priority of the module request is adjusted to obtain the adjusted priority. Based on the adjusted priorities of all functional modules, the module requests of each functional module are sorted to obtain the current set of execution strategies. Through a global strategy arbitration engine, the effective priority of each module is dynamically adjusted according to the risk value, which can automatically resolve request conflicts between functional modules and achieve globally optimal interactive control decisions.

[0055] For example, suppose there are module requests from multiple functional modules in the cockpit at the current moment, forming the current module request set M={m1,m2,…m_i…,m_n}, where m_i represents the module request issued by the i-th module. Each module request has an original priority P_i, which is weighted and adjusted using a risk sensitivity function g_i(R) to obtain the adjusted priority P_i′, expressed as: P_i′=P_i·g_i(R_t); Where m_i is the i-th module request, M is the current set of module requests, P_i is the original priority of the module request, P_i′ is the effective priority after continuous risk value adjustment; g_i(R) is the risk sensitivity function, which defines how the priority of the i-th module is adjusted as the risk value changes.

[0056] The global policy arbitration engine is responsible for sorting the requests from each module according to the adjusted priority P_i′, and generating the execution policy set C_t for the current moment, represented as: C_t=Sort(P1′,P2′,…,P_n′); Wherein, C_t is the set of execution strategies, including a list of module requests sorted by priority, which determines the execution rights and resource allocation of each functional module at the current moment.

[0057] Based on a risk-driven dynamic strategy arbitration mechanism, priority reconstruction and conflict resolution are carried out on multi-source interaction requests in the cockpit, thereby optimizing the human-computer interaction experience while ensuring driving safety.

[0058] In one embodiment, when the functional module is a security-related module, the risk sensitivity function corresponding to the functional module is a monotonically increasing function; when the functional module is a non-security-related module, the risk sensitivity function corresponding to the functional module is a monotonically decreasing function.

[0059] In this embodiment, the risk sensitivity function of a functional module can be set according to whether the module is related to safety. When the functional module is a safety-related module, the risk sensitivity function corresponding to the module is a monotonically increasing function; when the functional module is not a safety-related module, the risk sensitivity function corresponding to the module is a monotonically decreasing function. By setting different risk sensitivity functions for different functional modules, the optimal control method can be generated for the current cockpit state. Adding a new functional module only requires defining its risk sensitivity function g_i(R), without modifying the overall control architecture, which facilitates portability and adaptation between different vehicle platforms.

[0060] S150. Generate corresponding control instructions for each functional module based on the set of execution strategies.

[0061] In this embodiment, corresponding control instructions can be generated for each functional module according to the set of execution strategies, and the control instructions for different functional modules are different.

[0062] In one embodiment, generating corresponding control instructions for each functional module based on the execution strategy set includes: for each functional module in the execution strategy set, generating corresponding control instructions for the functional module based on the behavior strategy mapping function corresponding to the functional module.

[0063] The behavior policy mapping function can be a function that describes the relationship between module requests and corresponding control instructions. The behavior policy mapping function can be customized according to the characteristics of each functional module, and this embodiment does not limit it.

[0064] In this embodiment, the behavior strategy mapping functions of different functional modules are different. For each functional module in the execution strategy set, corresponding control instructions can be generated for the functional module based on the behavior strategy mapping function corresponding to that functional module.

[0065] For example, based on the execution strategy set C_t, corresponding control instructions U_i are generated for each functional module, represented as: U_i=Φ_i(C_t); Here, U_i represents the control instruction sent to the i-th module, which may include interface rendering complexity parameters, function enable or disable flags, and message notification scheduling strategies; Φ_i is the behavior strategy mapping function of the i-th module, which can map the set of execution strategies to specific module control parameters; C_t is the set of execution strategies at the current moment. After receiving the instruction, each functional module performs the corresponding interface switching, function restriction, or message notification management operation.

[0066] This invention provides a unified safety control method for an intelligent cockpit, comprising: acquiring multi-source sensing data from the intelligent cockpit; performing time alignment on the multi-source sensing data to obtain a time-series input sequence set; the time-series input sequence set including current time-series input data and historical time-series input data corresponding to the current time-series input data; selecting multiple time-series input data from the time-series input sequence set and inputting them into a time-series encoder to obtain a current state embedding vector; determining a current continuous risk value based on a pre-constructed risk mapping function and the current state embedding vector; sorting the module requests in the current module request set based on the current continuous risk value to obtain an execution strategy set; and generating corresponding control instructions for each functional module based on the execution strategy set. This method obtains the current continuous risk value by analyzing multi-source sensing data in the intelligent cockpit and determines how to control each functional module based on the risk value, thereby enabling unified control of the intelligent cockpit and solving the problem that existing cockpit control methods cannot adapt to dynamically changing driving environments.

[0067] Based on the above embodiments, modified embodiments of the above embodiments are proposed. It should be noted that, in order to keep the description brief, only the differences from the above embodiments are described in the modified embodiments.

[0068] In one embodiment, the method further includes: adjusting the current risk state when the current continuous risk value is continuously higher than the upper transition threshold and the duration exceeds the rising hysteresis time, or when the current continuous risk value is continuously lower than the backoff threshold and the duration exceeds the falling hysteresis time; and determining whether to retain module requests in the execution strategy set based on the adjusted risk state to obtain a filtered execution strategy set.

[0069] Among them, the upward transition threshold can refer to the threshold that the risk status needs to reach from a low level to a high level; the upward hysteresis time can refer to the time during which the continuous risk values ​​need to exceed the upward transition threshold when the risk status rises; the downward hysteresis threshold can refer to the threshold that needs to be reached when the risk status falls from a high level to a low level; and the downward hysteresis time can refer to the time during which the continuous risk values ​​need to be below the downward hysteresis threshold when the risk status falls. The upward transition threshold, upward hysteresis time, downward hysteresis threshold, and downward hysteresis time for different levels can be defined according to the actual situation.

[0070] In this embodiment, to avoid frequent switching of control commands due to short-term fluctuations in risk values, hysteresis control conditions are introduced. This embodiment defines an upward transition threshold T_up, a backoff threshold T_down, an upward hysteresis time τ_up, and a downward hysteresis time τ_down. Figure 4 This is a schematic diagram of risk status adjustment provided in an embodiment of the present invention, such as... Figure 4 As shown, taking low-risk, medium-risk, and high-risk states as an example, when the current continuous risk value R_t is consistently higher than the upper transition threshold T_up for a duration exceeding τ_up, the system triggers a transition to a higher-level state. When the continuous risk value R_t is consistently lower than the backoff threshold T_down for a duration exceeding τ_down, the system triggers a backoff to a lower-level state. This mechanism can be implemented using a state machine to ensure a smooth and stable control process.

[0071] Based on the adjusted risk status, it can be determined whether module requests in the execution strategy set should be retained, resulting in a filtered execution strategy set. For example, if the adjusted risk status is high-risk, some unnecessary functions should be prohibited from being called under high-risk conditions. In this case, the module requests corresponding to the functions prohibited under high-risk conditions can be removed. If they are not removed, the unnecessary functions simply have lower priority but may still be called. By further filtering the execution strategy set through the adjusted risk status, driving safety can be ensured.

[0072] In one embodiment, the method further includes: inputting the current state embedding vector and the historical state embedding vector into the prediction model to obtain the predicted risk value at a future time; and pre-adjusting the original priority of each functional module based on the predicted risk value.

[0073] The prediction model can be a pre-trained model for predicting future risk values, and the training method of the prediction model is not limited in this embodiment of the invention.

[0074] In this embodiment, a prediction model f̂(·) is introduced to estimate future risk values. The current state embedding vector and historical state embedding vector can be input into the prediction model to obtain the predicted risk value at future moments. Based on the predicted risk value, the original priorities of each functional module can be adjusted in advance. Through the prediction model, an upgrade from passive state response to proactive risk evolution control can be achieved, allowing for early intervention before risks occur.

[0075] For example, based on the current and historical state embedding vector sequence X_{tk:t}, the predicted risk value R_{t+Δ} for the future time t+Δ is expressed as: R_{t+Δ}=f'(X_{tk:t}); Here, R_{t+Δ} represents the predicted risk value at a future moment, f'(·) represents the prediction model, which can be obtained by joint training with the risk mapping function; Δ represents the prediction time span. The predicted risk value R_{t+Δ} is pre-input into the global policy arbitration engine to achieve prediction-driven control, and the priority of each module is adjusted in advance before the risk occurs.

[0076] Example 2 Figure 5 This is a schematic diagram of a unified safety control device for a smart cockpit provided in Embodiment 2 of the present invention. The device is applicable to situations where various modules in a smart cockpit are controlled in a unified manner. The device can be implemented by software and / or hardware and is generally integrated into a vehicle.

[0077] like Figure 5 As shown, the device includes: The acquisition module 210 is used to acquire multi-source sensing data in the intelligent cockpit, perform time alignment on the multi-source sensing data, and obtain a time-series input sequence set; the time-series input sequence set includes the current time-series input data and the historical time-series input data corresponding to the current time-series input data. Input module 220 is used to select multiple time-series input data from the time-series input sequence set and input them to the time-series encoder to obtain the current state embedding vector; The determination module 230 is used to determine the current continuous risk value based on a pre-constructed risk mapping function and the current state embedding vector; The sorting module 240 is used to sort the module requests in the current module request set based on the current continuous risk value to obtain an execution strategy set; The generation module 250 is used to generate corresponding control instructions for each functional module based on the set of execution strategies.

[0078] This embodiment provides a unified safety control device for an intelligent cockpit, comprising: an acquisition module for acquiring multi-source sensing data from the intelligent cockpit, performing time alignment on the multi-source sensing data to obtain a time-series input sequence set; the time-series input sequence set includes current time-series input data and historical time-series input data corresponding to the current time-series input data; an input module for selecting multiple time-series input data from the time-series input sequence set and inputting them into a time-series encoder to obtain a current state embedding vector; a determination module for determining a current continuous risk value based on a pre-constructed risk mapping function and the current state embedding vector; a sorting module for sorting the requests of each module in the current module request set based on the current continuous risk value to obtain an execution strategy set; and a generation module for generating corresponding control commands for each functional module based on the execution strategy set. By analyzing the multi-source sensing data in the intelligent cockpit to obtain the current continuous risk value, and determining how to control each functional module based on the risk value, unified control of the intelligent cockpit can be achieved, solving the problem that existing cockpit control methods cannot adapt to dynamically changing driving environments.

[0079] Furthermore, module 210 is specifically used for: Data from the smart cockpit is collected at a uniform sampling frequency through an onboard sensor network to obtain multi-source perception data; the multi-source perception data includes at least driver attention feature data, driver interaction behavior data, and vehicle motion dynamic data.

[0080] Furthermore, the input module 220 includes: A selection unit is used to define the length of a time window, and select the current time-series input data and historical time-series input data within the time window from the time-series input sequence set based on the time window length, as the current input sequence; The first input unit is used to input the current input sequence into the timing encoder to obtain the current state embedding vector.

[0081] Furthermore, module 230 is defined as including: The determining unit is configured to determine the rate of change of the current state embedding vector based on the current state embedding vector and the state embedding vector at the previous time step; The second input unit is used to input the current state embedding vector and the rate of change into the risk mapping function to obtain the current continuous risk value.

[0082] Furthermore, the sorting module 240 includes: The acquisition unit is used to acquire the current module request set in the intelligent cockpit at the current moment; the current module request set includes module requests from multiple functional modules and the original priority of each module request; An adjustment unit is used to adjust the original priority of the module request of each functional module based on the current continuous risk value and in combination with the risk sensitivity function corresponding to the functional module, so as to obtain the adjusted priority. Based on the adjusted priority, the module requests of each functional module are sorted to obtain the current set of execution strategies.

[0083] Furthermore, when the functional module is a safety-related module, the risk sensitivity function corresponding to the functional module is a monotonically increasing function; when the functional module is a non-safety-related module, the risk sensitivity function corresponding to the functional module is a monotonically decreasing function.

[0084] Furthermore, the generation module 250 is specifically used for: For each functional module in the execution strategy set, corresponding control instructions are generated for the functional module based on the behavior strategy mapping function corresponding to the functional module.

[0085] Furthermore, it also includes: When the current continuous risk value is continuously higher than the upper transition threshold and the duration exceeds the rising hysteresis time, or when the current continuous risk value is continuously lower than the fallback threshold and the duration exceeds the falling hysteresis time, the current risk status is adjusted. Based on the adjusted risk status, it is determined whether to retain module requests in the set of execution strategies, thus obtaining the filtered set of execution strategies.

[0086] Furthermore, it also includes: The current state embedding vector and the historical state embedding vector are input into the prediction model to obtain the predicted risk value for future time moments; The original priorities of each functional module are pre-adjusted based on the predicted risk value.

[0087] The aforementioned unified safety control device for intelligent cockpits can execute the unified safety control method for intelligent cockpits provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of executing the method.

[0088] Example 3 Figure 6A schematic diagram of a vehicle that can be used to implement embodiments of the present invention is shown. Vehicle 10 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The vehicle may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0089] like Figure 6 As shown, vehicle 10 includes at least one processor 11 and a memory communicatively connected to at least one processor 11. The memory can be read-only memory (ROM) and / or random access memory (RAM), such as ROM 12 and RAM 13. The memory stores computer programs executable by at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in ROM 12 or loaded into RAM 13 from storage unit 18. RAM 13 can also store various programs and data required for the operation of vehicle 10. The processor 11, ROM 12, and RAM 13 are interconnected via bus 14. An I / O interface 15 with input / output (I / O) functionality is also connected to bus 14.

[0090] Multiple components in vehicle 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows vehicle 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0091] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as the unified safety control method for smart cockpits.

[0092] In some embodiments, the unified safety control method for the smart cockpit can be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on vehicle 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the unified safety control method for the smart cockpit described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to perform the unified safety control method for the smart cockpit by any other suitable means (e.g., by means of firmware).

[0093] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0094] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0095] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory, read-only memory, erasable programmable read-only memory (EPROM) or flash memory, optical fibers, compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0096] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device for displaying information to the user, such as a cathode ray tube (CRT) or liquid crystal display (LCD) monitor; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0097] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0098] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system. It addresses the shortcomings of traditional physical hosts and Virtual Private Server (VPS) services, such as high management difficulty and weak business scalability.

[0099] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0100] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A unified safety control method for intelligent cockpits, characterized in that, The method includes: Acquire multi-source sensing data from the intelligent cockpit, perform time alignment on the multi-source sensing data, and obtain a time-series input sequence set; the time-series input sequence set includes the current time-series input data and the historical time-series input data corresponding to the current time-series input data. Multiple time-series input data are selected from the time-series input sequence set and input into the time-series encoder to obtain the current state embedding vector; The current continuous risk value is determined based on the pre-constructed risk mapping function and the current state embedding vector; Based on the current continuous risk value, the requests of each module in the current module request set are sorted to obtain the execution strategy set; Based on the set of execution strategies, corresponding control instructions are generated for each functional module.

2. The method according to claim 1, characterized in that, The acquisition of multi-source perception data in the intelligent cockpit includes: Data from the smart cockpit is collected at a uniform sampling frequency through an onboard sensor network to obtain multi-source perception data; the multi-source perception data includes at least driver attention feature data, driver interaction behavior data, and vehicle motion dynamic data.

3. The method according to claim 1, characterized in that, The step of selecting multiple time-series input data from the time-series input sequence set and inputting them into the time-series encoder to obtain the current state embedding vector includes: Define a time window length, and select the current time-series input data and historical time-series input data within the time window from the time-series input sequence set based on the time window length, as the current input sequence; The current input sequence is input into the timing encoder to obtain the current state embedding vector.

4. The method according to claim 1, characterized in that, The determination of the current continuous risk value based on the pre-constructed risk mapping function and the current state embedding vector includes: Based on the current state embedding vector and the previous state embedding vector, determine the rate of change of the current state embedding vector; The current state embedding vector and the rate of change are input into the risk mapping function to obtain the current continuous risk value.

5. The method according to claim 1, characterized in that, The execution strategy set is obtained by sorting the module requests in the current module request set based on the current continuous risk value, including: Obtain the current module request set within the intelligent cockpit at the current moment; the current module request set includes module requests from multiple functional modules and the original priority of each module request; For each functional module, based on the current continuous risk value and combined with the risk sensitivity function corresponding to the functional module, the original priority of the module request of the functional module is adjusted to obtain the adjusted priority; Based on the adjusted priority, the module requests of each functional module are sorted to obtain the current set of execution strategies.

6. The method according to claim 5, characterized in that, When the functional module is a safety-related module, the risk sensitivity function corresponding to the functional module is a monotonically increasing function; when the functional module is a non-safety-related module, the risk sensitivity function corresponding to the functional module is a monotonically decreasing function.

7. The method according to claim 1, characterized in that, The generation of corresponding control instructions for each functional module based on the set of execution strategies includes: For each functional module in the execution strategy set, corresponding control instructions are generated for the functional module based on the behavior strategy mapping function corresponding to the functional module.

8. The method according to claim 1, characterized in that, Also includes: When the current continuous risk value is continuously higher than the upper transition threshold and the duration exceeds the rising hysteresis time, or when the current continuous risk value is continuously lower than the fallback threshold and the duration exceeds the falling hysteresis time, the current risk status is adjusted. Based on the adjusted risk status, it is determined whether to retain module requests in the set of execution strategies, thus obtaining the filtered set of execution strategies.

9. The method according to claim 1, characterized in that, Also includes: The current state embedding vector and the historical state embedding vector are input into the prediction model to obtain the predicted risk value for future time moments; The original priorities of each functional module are pre-adjusted based on the predicted risk value.

10. A unified safety control device for an intelligent cockpit, characterized in that, The device includes: The acquisition module is used to acquire multi-source sensing data in the intelligent cockpit, perform time alignment on the multi-source sensing data, and obtain a time-series input sequence set; the time-series input sequence set includes the current time-series input data and the historical time-series input data corresponding to the current time-series input data. The input module is used to select multiple time-series input data from the time-series input sequence set and input them into the time-series encoder to obtain the current state embedding vector; The determination module is used to determine the current continuous risk value based on a pre-constructed risk mapping function and the current state embedding vector; The sorting module is used to sort the module requests in the current module request set based on the current continuous risk value to obtain an execution strategy set; The generation module is used to generate corresponding control instructions for each functional module based on the set of execution strategies.

11. A vehicle, characterized in that, The vehicles include: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the unified safety control method for intelligent cockpits as described in any one of claims 1-9.

12. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the unified safety control method for intelligent cockpits as described in any one of claims 1-9.

Images