Mitsubishi Electric R&D Centre Europe

The disclosure relates to a computer program verification method, the computer program verification method comprising,- a computer program transformation phase (P01) in which an input program (InPrg) that comprises at least one function call to at least one abstract mathematical function (f_i) is transformed in *an output program (OutPrg), the output program (OutPrg) comprising at least one call to a lower precision implementation f̂_i of the at least one mathematical function f_i and at least one error verification directives (evd) linked to the lower precision implementation f̂_i of the at least one mathematical function f_i,- a computer program verification phase (P02) in which the output program (OutPrg) is provided to a formal verification module (FVM), the formal verification module (FVM) using the error verification directives (evd) to ensure correctness (cor) of the input program (InPrg).

A monitoring method of a gate oxide degradation in a power semi-conductor module comprising:- injecting successively two opposite currents into the gate and obtaining a flatband voltage difference ΔV_FB between the two gate-to-source voltage shifts of the two injections;- deducing a degradation level of the gate oxide of the said module.

A reconfigurable converter (100, 101) has a transformer stage (40) of matrix type, with several controlled rectifier units (51, 52,...) dedicated to separate transformer assemblies (41, 42,...). The reconfigurable converter is configured for inhibiting at least one of the rectifier units in a partial operation mode, so as to vary an effective turn ratio value of said reconfigurable converter. The inhibition can be produced through appropriate control of the rectifier units, without additional controlled switches being necessary. The reconfigurable converter may form a battery charger for transferring electrical power from an AC-grid to a battery set. Transition to the partial operation mode allows using components with lower breakdown limit and meeting thermal constraints for these components.

A process for collecting information by a central unit (100) from at least one remote unit (1) comprises transmitting at least one of a characterization of a proposed available information (CH(PAI)) and a synthesis of an already-collected information (SYN(ACI)) before at least part of the available information (SEL_INFO) is sent by the remote unit to the central unit. The process allows reducing a transmission payload by deleting overlapping information, or prioritizing information overlaps for providing increased reliability, or also focusing sent information on a target. The process may be used for applications that involve exploration of an environment and building a digital representation of a content of the environment, by collecting sensing results obtained by the at least one remote unit.

Device for sensing a junction temperature of a power semiconductor (1) controlled by a PWM generator (3), comprising a pull-down resistor (10) between a gate and an emitter of said power semiconductor (1), a sense switch (11) connected between said emitter and a ground connection, a command circuit (12) for turning ON and OFF of said sense switch, said command circuit (12) being triggered by said PWM generator (3), a measurement means (13) configured to monitor a voltage response V<E> between said emitter and ground during OFF periods of said sense switch and conversion means to calculate an internal gate resistance R<gin> of the power semiconductor representative of said junction temperature based on said voltage response V<E>.

A process for collecting information by a central unit (100) from at least one remote unit (1, 2) comprises determining a selection to be applied to information available at the remote unit and applying the determined selection to the available information by said remote unit. The selection may be determined either by the central unit or at each remote unit, depending on alternative implementations. The process allows exploiting redundancy for various purposes, including reducing a radiocommunication payload, or providing increased reliability, or also focusing sent information on a target. The process may be used for applications that involve exploration of an environment and building a digital representation of a content of the environment, by collecting sensing results obtained by the at least one remote unit.

Method (500) for jointly transporting a plurality of data flows via a system (2) having a limited capacity, the plurality of data flows being respectively associated to sets of requirements, the method comprising the following steps: - (S510) obtaining information related to the limited capacity and to the sets of requirements, - based on said plurality of data flows, on the information related to the limited capacity and the sets of requirements, (S540) selecting at least one aggregation mode for aggregating at least a part of the plurality of data flows, among a plurality of predefined aggregation modes, - (S550) applying said selected aggregation mode to said part of the plurality of data flows to determine an aggregated data flow, - (S560) encoding the aggregated data flow using an encoding function, - (S570) transporting the encoded aggregated data flow via the system (2).

Laser welding method for welding a first copper layer (141, 200) on a second copper layer (140, 301) with a laser beam (110) emitted by a laser welding station from above the first copper layer across a convergent lens (120), wherein height (H) of the convergent lens above said first copper layer (141) is adapted to provide a focal point (F) of said laser beam above the first copper layer in order to provide at least one defocused welding spot (S) on said first copper layer.

The invention relates to a network interface controller (NIC) sharing method, the method being implemented by a virtual machine (RTOS VM) running Real-Time Operating Systems (RTOS) in charge of executing time critical tasks, the method comprising: - implementing (S01), by the RTOS, a hardware driver (NICd) for exclusively driving the NIC from the RTOS, - implementing (S02), by the RTOS, a backend driver (BE-vNIC) for driving a virtual NIC (vNIC) from said RTOS, the backend driver (BE-vNIC) enabling receiving and transmitting network frames to at least one user virtual machine, each of the at least one user virtual machine using an at least one corresponding frontend driver, - the RTOS setting (S03) at least one shared memory region (SMR) of a memory space dedicated to the RTOS VM, to store and exchange the network frames of the at least one user virtual machine for enabling the backend driver (BE-vNIC) to exchange the network frames with the at least one frontend driver.

A process for transmitting data between two communication units comprises selecting a transmitter coding scheme based on at least two communication criteria assessed for several modulation and coding schemes. The communication criteria may be assessed using an interference model, and the process is compatible with implementing superposition coding. The at least two communication criteria may be an expected rate and a variance or standard deviation of an instantaneous rate of the data transmission. The process allows selecting the transmitter coding scheme such that it suits the interference model according to a trade-off rule applied to the communication criteria.